Oikonomeo

Why do we tax capital gains at a lower rate?

2012-01-24T17:10:00.002-05:00

Mitt Romney paid about 14% of his income in taxes the last 2 years. This is for a number of reasons, but mainly it's because most of his income was classified as capital gains.

Why do we tax capital gains at a lower rate? There actually is a good economic rationale: Taxing returns on investment is effectively taxing income twice.

Suppose that you receive $80 pre-tax, and you can either choose to spend it now, or to save it for 10 years at the end of which it will earn a return of 50%.

If there were no taxes, this is a choice between $80 of consumption today, or $120 in 10 years.

If there were income taxes of (say) 50%, then after taxes you would have $40. This could be spent on $40 of consumption immediately, or saved to earn an additional $20 in interest. But since those interest payments are income, they would be taxed at 50% and so 10 years from now you could increase your consumption by $40 + $10 = $50.

Thus an income tax rate of 50% implies an effective tax rate of 1/2 on immediate consumption, or of 7/12 on delayed consumption. Since consumption is what people care about when making decisions, income taxes provide an incentive to consume immediately rather than to save.

Of course, the best solution would be a pure consumption tax system, but we get some improvement by having lower tax rates on income from interest payments on savings, which is why we tax capital gains at a lower rate.

Still a Long Way to Go

2012-01-13T16:01:00.005-05:00

Greg Mankiw suggests that the liquidity trap may soon be over. Paul Krugman disagrees. So just how far have we come since the recession ended in the middle of 2009? In my view we still have a long way to go. Here are a few (somewhat depressing) pictures to illustrate this point.

The first graph is investment as a share of GDP. Typically, investment falls more than other components of GDP during a recession. During normal times, investment is around 16% of GDP. While investment has improved a little since we hit bottom, it has stalled out at around 12.5% for the last 1.5 years. To put that into perspective, 12.5% would be the lowest share of GDP since WW2 if we excluded 2008-2009.

The second graph is the employment to population ratio. This is the fraction of the adult population that is employed. As you can see, it has essentially not recovered at all since we hit bottom. Since in the decade preceding the recession there was a slight downward trend, we may be up a little since the recession relative to trend, but we still have a long way to go.

Macroprudential Regulation

2011-11-15T18:00:00.002-05:00

One source of market failure in economic theory is when a market transaction between two agents in the economy imposes some cost on another agent. This cost is called an externality. Since the parties to the transaction do not experience this cost directly, it is not reflected in the price of the good. Therefore goods with negative externalities will be oversupplied by an unregulated market. The classic example of this is pollution.

Another example of an externality is systemic risk. When a financial institution takes on debt, it increases the risk of financial crisis, but this will not be reflected in market prices. Just as in the case of pollution, the solution is to tax the activity so that firms internalize this cost. That is macroprudential regulation.

The beginnings of the African Miracle?

2011-11-04T14:35:00.002-04:00

I should point out that the information quoted in my last post is a little out of date:

...over the ten years to 2010, six of the world’s ten fastest-growing economies were in sub-Saharan Africa.... Over the past decade the simple unweighted average of countries' growth rates was virtually identical in Africa and Asia. Over the next five years Africa is likely to take the lead. In other words, the average African economy will outpace its Asian counterpart.

(Source)

An essentially African phenomenon

2011-11-04T14:28:00.003-04:00

Xavier Sala-i-Martin (2005):

This disappointing performance [of African economies], together with the great success of the other two poor regions of the world (East and South Asia) means that the majority of the world's poor now live in Africa. Indeed, Africa accounted for only 14.5 percent of the world's poor in 1970. Today, despite the fact that Africa accounts for only 10 percent of the world population, it accounts for 67.8 percent of the world's poor (see the bottom panel of Table II). Poverty, once an essentially Asian phenomenon, has become an essentially African phenomenon.

Inequality Revolution

2011-11-03T18:21:00.002-04:00

The biggest increase in global income inequality happened during the Industrial Revolution.

The REAL 99%

2011-11-03T11:00:00.003-04:00

Global inequality has fallen substantially over the past 30 years, and more than over any comparable time frame in human history.

Can Republicans be Keynesians?

2011-09-01T10:13:00.004-04:00

Is there anything inconsistent about being a Republican and being a Keynesian?

First let's define our terms. I define a Republican as someone who believes that overall government interference in the economy should be reduced at the margin. If we use government spending as a proxy for total government intervention, this means that Republicans want lower spending at the margin. Since spending needs to be financed, this also means lower taxes at the margin.

A Keynesian is someone who believes that government spending should be counter-cyclical. A counter-cyclical variable is one that is relatively high during recessions, and relatively low during expansions. That is, it moves counter to the cycle.

From these definitions we can see that, since Republicanism is defined in terms of the average LEVEL of government spending and intervention, whereas Keynesianism is defined in terms of the cyclical TIMING of this spending, there is in principle no contradiction between the two. A Keynesian Republican would argue that government spending should be higher during recessions than during expansions, but should be lower overall. A non-Keynesian non-Republican would argue that government spending should be no smaller, but should be constant over the business cycle.

Of course, in practice if one were a Keynesian Republican, one's policy preferences relative to current policy would be to increase spending (Keynesian) and to decrease spending and decrease taxes (Republican), with the net policy being to cut taxes, and move spending little if at all.

So maybe the better judge of one's Keynesianism is how worried one is about the deficit during a recession, and the best judge of one's Republicanism is whether one prefers expansionary activity to happen through spending increases or tax cuts.

By this definition, I am a Keynesian Republican. Good to have that settled.

How to interpret stock and bond prices

2011-08-15T17:45:00.002-04:00

A quick guide to interpreting the markets, with an application to the news from the last week:

Finance 101: The fundamental value of an asset is the discounted value of future income flows from that asset. So when the stock market goes up, this indicates that investors expect more income sooner from these assets. Since these assets are claims on corporate profits, this means that they expect corporate profits to rise. Since corporate profits tend to rise during expansions and fall during contractions (recessions), a rise in expected corporate profits is good news for the economy. So the stock market going up is good news, and the stock market going down is bad news.

So what determines bond prices? Bonds are different than stocks in two respects. First, they are (generally) nominal assets, meaning that they are in the currency of a particular government. Secondly, they pay a fixed predetermined amount, unless the holder declares bankruptcy and defaults on them. Therefore they are in general much less risky.

The risk in bond prices is of two types. There is inflation risk -- if the currency in which they are denominated experiences inflation, they will lose value and yield a real return below their nominal yield. There is also default risk -- if the issuer runs into financial difficulty it may not repay.

Therefore a rise in bond yields (which is a fall in price) can mean three things. First, it can mean that inflation risk has gone up, and investors demand a higher nominal return. Secondly, it can mean that default risk has gone up, and so investors require a higher yield to compensate them for the risk that the issuer will not repay. Third, it can mean that the return from other investments has risen, and therefore investors require a higher return in order to accept bonds (this is arbitrage).

So how do we interpret stock and bond movements? Well, if we get good news about the economy, we should expect stock prices to go up. Since inflation tends to accompany expansion, both the inflation risk and rising returns on alternatives should raise bond yields. This shouldn't substantially affect default risk, so we should generally expect bond yields to go up too. If we get bad news, by the same logic we should see stock prices and bond yields fall.

On the other hand, suppose we find out that a country or firm is in bad fiscal shape for reasons apart from economic fundamentals. Then bond yields will rise because investors believe that a default is more likely, or that the country may run a higher inflation rate in order to reduce the real burden of its debt. Either of these should drive up bond yields, but since inflation or a default would likely hurt the economy, we should see stock prices fall.

In other words, when stock prices and bond yields move together, this is because of economic news (whether good or bad). When stock prices and bond yields move in opposite directions, this is a sign of fears of default or inflation apart from economic fundamentals.

Over the last week, the stock market plummeted. This was interpreted by some observers as a consequence of S&P's rating downgrade. However, the fact that bond yields fell as well, and especially that the US bond yield fell relative to other countries' bonds, indicates that what we saw in the last week was the result of bad news about economic prospects, not concerns about the safety of US debt.

If anything, all the signs indicate that investors really really want to hold our debt (the so-called 'flight to quality'). There's a school of thought that says that it is wise to make hay while the sun shines. If that's the case, there's no better time to borrow and spend than when the US government can borrow for 10 years at real interest rates below zero.

The Rating Downgrade: A Natural Experiment in Finance Theory

2011-08-06T11:02:00.003-04:00

Update: 10-year yields after one week: 2.25 (US), 2.33 (Germany), and 2.53 (UK). So now US yields will have to go up 90-100 basis points relative to Germany and UK to achieve Terry Belton's prediction.

*****************************************

In my previous post I laid out the likely consequence of a downgrade of US debt under several views of how pricing works. Now that S&P has downgraded us, we have a natural experiment.

Per Bloomberg, JPMorgan’s Terry Belton says that a rating downgrade could raise US treasury yields by 60 or 70 basis points. If this is true, we should observe this over the next several days. But how do we distinguish between the effect of the rating downgrade and movements due to economic fundamentals?

There are basically two things that determine the yield on government debt: perceived risk and the availability of good alternatives. Thus a rise in yield could be good (meaning the economy is recovering) or bad (meaning the bond's perceived riskiness has increased). So if we want to attribute changes in US bond yields to changing perception of risk, we need some way of subtracting changes due to available alternatives.

The way we usually do this is by looking at spreads. So we can look at the spread between yields on US treasuries (say 10-year bonds), and those of some safe European countries, say the UK and Germany.

At the close of Friday, those yields were 2.69% and 2.35% , compared to 2.56% for the US (source). So the US was right in the middle of the pack, about 10 basis points below the UK and 20 points above Germany. If the US adds 60-70 points, we would be 50-60 points above the UK and 80-90 points above Germany. So that's a rough baseline.

Let's see where we are in a week.

What would be the effect of a rating downgrade?

2011-08-01T20:39:00.003-04:00

Recently there has been some talk of a rating downgrade on US debt. Suppose this happened, and the rating agencies downgraded our debt from AAA to AA. What would be the effect of this downgrade?

The answer depends on your model for asset prices.

1. Fundamental View

Sure, market efficiency has taken some hits lately, but it's still economists' go-to model for thinking about a new question about asset markets. In this view, market participants use all available information to rationally judge the probability of default, and set interest rates accordingly (i.e. they value the asset based on its fundamentals). Under this view, a rating downgrade based on publicly available information would have no effect on interest rates whatsoever. S&P's views don't give investors any new information, so it doesn't effect what they require to hold government debt.

2. Sunspot View

There may be multiple rationally supportable equilibrium interest rates, and the rating agencies play a coordinating view. This stems from the view that high interest rates make debt more difficult to service, and therefore raise the probability of a default. Therefore we might have two (or more!) equilibria: a "good" equilibrium where everyone thinks that US debt is safe, so everyone charges low interest rates, the US is able to service its debt at low rates, and therefore US debt is safe; and a "bad" equilibrium where everyone thinks that US debt is risky, therefore everyone charges a high interest rate, therefore the US as trouble servicing its debt, and a future default becomes more likely (self-fulfilling prophecy).

What determines what equilibria we wind up at? Either one is rational, but only if all the investors are coordinated. You can coordinate around some external information, even if it's not directly relevant to the fundamentals of the asset at all (i.e. sunspots). Perhaps rating agency ratings serve such a role. In this scenario, a rating downgrade might bump us into a bad equilibrium right away and raise rates, or it might merely make a future run on US debt more likely.

3. Irrational or Bounded Rationality View

In this view, some investors behave irrationally (i.e. follow the rule "only buy securities S&P rates AAA"), and so would take a ratings downgrade into account. This could be because these investors are not very smart and make mistakes. It could also be a rational response to recognizing your own limitations at accurately synthesizing all publically available data. That is, perhaps gathering information and figuring out the right investment strategy is not free but requires the expenditure of resources (time and effort). Then it makes sense to "outsource" some of this decision-making by following information-rich signals from other investors. A third possibility is that there is a principal-agent problem, and some investment funds don't want their fund managers to take self-serving risks with their money, and so they bind their hands by having rules like "X% of the portfolio must be composed of assets rated AAA by S&P."

Of course, even if you have some irrational traders in the market, as long as there aren't too many the rational fundamental traders could in theory make up for them. When the investment funds "irrationally" sold the US debt they hold and want to buy AAA assets, other rational investors sell them AAA assets and buy up the government debt, recognizing that S&P's downgrade makes no practical difference.

Conclusion:

Okay, so which of these views is true? I would argue that #1 is a pretty good first approximation, but it can't be strictly true (since if it were, the rating agencies would not exist in the first place). #2 is probably true for some debt at some times, but probably not for the US right now. #3 definitely holds for some investors and would result in a small rise in interest rates due to a downgrade, but it would be hardly noticeable.

So my prediction is that a rating downgrade would make almost no difference in interest rates on US debt (which by the way are currently incredibly low -- seems the markets trust us even if S&P does not). There would be some movement since many funds do have rules about holding X% of AAA assets, but these movements would be limited and might be entirely temporary.

And for the record, Krugman points out that when Japan was downgraded it made no noticeable difference. Score one for the market fundamentalists!

What would be the result of a default?

2011-07-29T20:30:00.000-04:00

Suppose that the US Congress were to fail to raise the debt limit. What would be the consequence?

Let's be clear what we're talking about. I'm thinking of a failure to pay some fraction of holders of US treasuries, or US debt. I'm not thinking of a full-scale default (that is, all treasuries are worthless), but just that some treasuries are not paid in full.

The immediate effect would be that treasuries would be considered more risky. Demand for treasuries would fall as investors who held treasuries tried to sell them and put their money in other forms. This would cause interest rates to rise.

What would be the consequence of this? It's important to understand that US debt is not just any asset. It's considered a very safe asset, pretty much the safest asset there is. Now safe assets play a particular role in the economic system. They have high "liquidity", which means that their value can be accessed in a hurry, and in all states of the world. Companies, especially in the financial sector, like to keep a good bit of their net worth in very liquid assets so that they don't run into short-run financing problems which might force them to sell less liquid assets at a loss.

If something happens that reduces the supply of safe assets in the economy, this leads to a reduction in liquidity and what's commonly called a "liquidity crisis". We had one of those in Fall 2008, when a bunch of assets (basically assets based on slices of mortgages) that had been perceived as quite safe turned out to be very risky. The total value of those assets was quite small relative to the value of US debt in the world economy right now. Thus a default on treasuries would represent a much much larger decrease in the supply of liquidity in the world economy. This would likely precipitate a financial crisis many times the size of the one we saw in Fall 2008.

Guide to the Debt Ceiling Debate of 2011

2011-07-29T09:33:00.004-04:00

Okay, let's review the facts.

1. Fiscal situation

- In the late '90s, we temporarily had a budget surplus. This was aided by the tech bubble, but a small deficit is actually not a big deal. Economically, what matters is debt-to-GDP ratio. So fundamentally, our country had sound short-run finances as recently as a decade ago.

- From 1999 to 2007, the debt went up a fair bit, mostly due to paying for wars in Iraq and Afghanistan, and paying for tax cuts. Neither of these significantly alter the fiscal outlook going forward.

- Starting in 2007, we had a nasty recession cum financial crisis, and today we find ourselves with higher debt levels and running large deficits.

- Most of the increased debt is due to falling tax revenues (due to the recession) and higher expenditures on automatic stabilizers, bailing out the financial system, plus a little bit in stimulus.

- Unless we have a decade-long depression, this is a temporary problem, and doesn't seriously damage our fiscal standing.

- In the long-run, we have a fiscal problem: starting around a decade in the future, we are projected to run increasingly large deficits. This is almost entirely due to projected increases in medical costs, plus some demographic shifts.

Conclusion: Our fiscal problem is mostly a long-run problem. In the short-run, our chief concern should be restoring full employment. The best policy would likely be increased spending and monetary expansion designed to boost growth, combined with a credible commitment (i.e. a plan, preferably passed into law) to restore fiscal balance and deal with the long-run budget problem once the recession is over.

2. The debt ceiling

- Congress has a self-imposed limit on how much debt the Treasury can issue. Since Congress also passes tax and spending programs, the debt limit's only function is to make defaulting on our debt the default option without additional Congressional action. Yet somehow US debt is considered the safest asset in the world.

- Traditionally, the debt limit has served a purely political role: the minority party is allowed to gain some political capital for itself by voting against raising it, and talking about how much they dislike debt. I'm not sure why voting in favor of defaulting on our debt is considered "good" -- I would think that most Americans are okay with taking out a mortgage, and would consider it wrong to stop paying it on the grounds that 'debt is bad.' One of life's little mysteries I suppose.

- We can see this political role at work in, for instance, the fact that Barack Obama voted against raising the debt limit in 2006 (i.e. voted in favor of defaulting on our nation's obligations) when he was in the Senate.

- Now that we have split government, Democrats needed Republican votes to raise the debt limit. Republicans believed that they should receive some concessions in return for these votes, since this has traditionally been a source of political capital for the minority party.

- However, since Republicans think that debt is REALLY REALLY bad (presumably they oppose on principle mortgages, car loans, student loans, and small business loans), they think that an appropriate price for their votes is to solve our long-run fiscal problem AND balance the budget in the short-run, and to do so without any increase in taxes.

- Obama, fancying himself a somewhat conservative moderate, decided to craft a grand bargain, which conceded the Republican's premise, but insisted that maybe a little bit of revenue should be raised via new taxes. This is approximately the sort of deal that someone like Romney or Huntsman would pass if they had control of both houses of Congress.

- No deal.

Conclusion: The Republican party appears to be crazy.

3. Out current status

- After much debate, there are now two bills: one in the House (Republican) and one in the Senate (Democrat).

- They are the same in almost all particulars. They are a perfect Washington compromise: they give into Republican demands (no new taxes) and Democrat demands (no substantial reforms of Social Security or Medicare). In other words, they don't really accomplish anything, except for slashing some spending in the short run (which will hurt the economy).

- They differ in one major respect: the Democratic bill raises the debt limit by more, so that we won't need to raise it again until after the 2012 elections. The Republican bill raises it by less, so that we will need to raise it again before the 2012 elections.

- In other words, the bills accomplish nothing good, potentially do some harm, and the current hold up on determining whether the US will default on its promises is entirely a matter of political advantage in the next round of ridiculousness.

Conclusion: These people run the world?

Confusion of Positive and Normative

2011-06-20T11:16:00.004-04:00

I am often struck by how people often conflate normative and positive statements. What I mean is confusion between "X should be true" and "X is true". I see this most often in politics, and it frustrates and baffles me.

A few examples will illustrate the fallacy. Consider the Second Amendment to the Constitution: "A well regulated Militia, being necessary to the security of a free State, the right of the people to keep and bear Arms, shall not be infringed." Now consider the positive question, "Does the Second Amendment prohibit gun control laws?" and the normative question, "Should we pass gun control laws?"

I consider these questions to be separate. One can consistently hold the views, "Guns should be regulated," and "Gun regulation is unconstitutional." This view would only be inconsistent if one also held the view, "Anything that is unconstitutional should not be done." [After modifying the constitution, of course].

Another example is the positive legal question, "Was Roe v. Wade wrongly decided?" and the normative questions, "Should women have the right to have abortions?" These are different questions, yet empirically it seems that people hold beliefs that correspond nearly one-for-one.

A third example, just to show I'm not making a point about the constitution: consider the questions, "Should we have invaded Iraq?" and "Did the US invasion of Iraq play a role in bringing about the Arab Spring?" It is entirely logically consistent to believe that we should not have invaded Iraq and that the invasion played a role in the Arab Spring. It is likewise consistent to believe that the US invasion played no role, but that we were nevertheless justified in invading Iraq for other reasons. Yet in practice such combinations are very rare.

I should note that while it's logically consistent to hold "opposite" views on such questions, there are sometimes good reasons that they go together. For instance, if you think that the Constitution is generally right, then the constitutionality of something might be a good normative guide. If you think that the Iraq War was justified largely because it increased the likelihood of an Arab Spring scenario, then the Arab Spring could provide evidence to justify the invasion. Yet it seems to me that the logical connections between such beliefs are much weaker than their empirical regularity.

I think that this confusion between the normative and the positive is a key part of the problem with our national debate. The purpose of logic is to clarify the relationships between propositions. The goal of debate is to isolate the source of disagreement through the application of logic. Then when we have cleared away our confusion, we can examine any remaining difference of belief in light of the evidence.

When we confuse the positive and the normative, logical debate loses this clarifying role. Any point that bears on an issue of national interest will invariably support one party's preferred policies over the other's. Debate becomes a zero-sum game and no agreement can be reached on purely logical points.

This is a terrible state of affairs.

Policy versus Traps

2011-03-18T17:14:00.002-04:00

Perhaps one of the most surprising things about economic growth is how unusual it is historically. Before 1800, consistent economic growth was essentially unheard of anywhere in the world. From 1800-1900, economic growth was restricted to a few countries in Western Europe and offshoots. It's only in the past 30 years that a majority of the world's population have lived in countries that have experienced sustained economic growth.

Given these facts, there's a natural tendency to move from viewing growth as a natural state of an economy to something rare and mysterious. Yet the reality is that the growth rate for countries experiencing economic growth has been fairly stable. Countries near the technology frontier have enjoyed per-capita real growth on the order of 2% per year. This implies that per-capita output doubles approximately every 35 years. Countries below the technology frontier are able to grow quite a bit faster than this, with their upper potential growth rates increasing proportionally with how far they are below the technology frontier, although many poorer countries remain in no-growth or low-growth states.

Given this remarkable consistency, one wonders why the world went so long without enjoying economic growth. If growth of around 2% per year is possible with a fairly simple set of institutions, then why did it take so long for anyone to unlock its potential?

One potential answer may seem quite surprising -- perhaps economic policy was just really bad in all countries until two centuries ago, and in most countries until the last several decades. This may seem a surprising idea -- that we so consistently were shooting ourselves in the foot for so much of human history. Yet when you look at the countries that have experienced economic growth, you notice a remarkable consistency. Economic growth invariably corresponds to a basic level of openness, freedom, and stability. Where this is absent, so is growth. Where it is present, economic growth is as well.

Of course, this may be reverse causality at work. Perhaps there is a "poverty trap" argument for economic growth, in which poor countries require some large shock to "jump start" the growth engine. Perhaps once started growth inevitably changes the culture and empowers the masses, so that freedom and openness follow. Disentangling these alternative explanations is a difficult task, and as is so often the case they have very different policy implications.

Supply versus Demand

2011-01-28T23:16:00.004-05:00

A quick add-on to my last post.

This plot shows GDP (blue), which is what the economy produced. It shows (annualized) percentage growth each quarter. The red-orange line shows demand. Anything that is produced and not sold (excess supply) goes into inventories, so demand is just output (GDP) minus changes in inventories. The yellow line shows the rate of growth that the economy has to achieve to keep pace with population and productivity growth (around 2.3%).

What we see is that output began expanding in the second quarter of 2009, reached above the water line (i.e. firms started hiring) in late 2009 and early 2010, but then dipped below the water line again in the Spring/Summer, and is only now rising ever so slightly above it. Not so encouraging.

When we consider inventories we get a different story. Demand didn't start growing again until late 2009, and even then never cleared the water line. The increase in output in late 2009 and early 2010 outpaced demand, which may have caused the slowdown in output of the past several quarters.

Which brings us to right now and the spike in demand in the fourth quarter of 2010 that I talked about in my last post.

Incidentally, James Hamilton comes to somewhat different conclusion from me. He argues that inventories and imports are substitutes in the data. This would imply that some of the excess inventory accumulation of the past year is being sold at reduced prices and consumers are buying those goods in place of imported goods.

However I doubt this can explain all of the spike in demand we observe. Even if the whole fall in imports (2.4%) is assumed to have come out of inventories of imported goods, that still leaves us with around 5% growth in domestic demand, which is very different from anything we have observed so far in this recovery. The latest report may be nothing special on the output side, but I think that it's the best news we've gotten so far on the demand side.

A prediction about GDP

2011-01-28T22:31:00.004-05:00

The early Q4 numbers are out, and they say we grew at 3.2% last quarter.* That doesn't seem too impressive. Paul Krugman points out that at this rate we'll be back at full employment in 2018. Brad DeLong says that at least it's better than being poked in the eye with a stick.

They're right about the headline figure. GDP grows at about 2% just based on population and productivity growth. Increasing the labor force by 1% boosts GDP by around 2%, so 3.2% growth over a year corresponds to shaving a little more than half a point off the unemployment rate (which is right now at 9.4%).

Yet when you look beneath the headline figure, you see a much rosier picture. That 3.2% is composed of a number of components: Consumption, investment, government spending, and net exports. These refer to the demand side of the economy, but we're clearly demand-constrained right now (high unemployment plus deflation).

So what do we see? Consumption is up 3.04% of GDP, investment is down -3.20%, net exports are up 3.44%, and government expenditures are down -0.11%. 3.04 - 3.2 + 3.44 - 0.11 = 3.17, which is where we get our 3.2% figure. So the increase is being driven by large increases in demand (at home and abroad).

Why did private investment fall by -3.2%? This is the exciting part. You see, in GDP accounting we include changes in inventories in investment. This is really just an accounting point. GDP is supposed to measure current period production, so if Ford produces some cars that it doesn't sell, we want to count those in the quarter in which they are produced, not in which they are sold. So they get added to inventory.

It turns out that this -3.2% figure is -3.7% inventories, and +0.5% everything else. In other words, while we only produced 3.2%* more than we did last quarter, demand actually increased by closer to 7%.

What does this mean? Well as I said before, the economy is currently demand-constrained, not supply-constrained. This huge drop in inventories can't continue, since firms only have so much in inventory. So what will firms do in the face of rising demand? Increase capacity utilization, meaning a fall in unemployment and a boost to investment. And yes, that will show up in the headline figure.

So here's my prediction: 2011 will see major growth. I'm not going to predict 7% growth, but I do expect 4-5% growth. And at 5% growth, we'll be back to full-employment by the end of 2013.

* Quarterly GDP is reported in annualized numbers, so 3.2% growth means we were up about 0.8% last quarter, which if continued over a year would be 3.2%.

Conservatism, economics, and the insanity defense

2011-01-21T11:28:00.011-05:00

Following up on my last post about the Tucson shooting, today I was reading the New York Times Room for Debate about the insanity defense. It turns out that, after Hinckley was found not-guilty by reason of insanity for shooting Ronald Reagan, many states rewrote the standards for the insanity defense. It used to be that it was sufficient to demonstrate that someone was mentally ill and was therefore not responsible for their actions. Now the burden is placed on the defense[1] to "demonstrate that he did not appreciate the wrongfulness of his conduct."

Whether you think this is a good or bad thing depends on your views of the purpose of the criminal justice system. If you think that the purpose of the criminal justice system is to do justice, meaning to punish those who commit moral crimes and to exonerate those who do not, and that punishments should be proportional to harm intended and done, then this seems to be a reasonable standard. What matters is whether somebody was morally culpable.

As an economist, I view things somewhat differently. The purpose of the criminal justice system is to reduce crime. One major way it does this is by punishing people who are caught doing certain activities that we want less of, therefore creating incentives for people not to engage in those activities. From this perspective, punishing an insane person is ridiculous, since this will not affect the decisions of future insane people. If the goal is the pragmatic reduction of crime, then insane defendants should receive treatment and should be kept somewhere where they cannot harm others unless and until they recover and are able to re-enter society.

This debate reminds me of a correspondence I had with a friend recently. He asked me for my views on an article about the moral case for conservative economics. From my response:

Economic logic implies a number of principles, many of which are in line with what are broadly considered to be "conservative" principles. However, this correspondence is mostly coincidental, and does not hold consistently.

For example, consider the principle of personal responsibility. Loosely speaking, this is the principle that outcomes should be closely tied to one's actions. Political conservatives (in the US) consider this a fairly bedrock principle, and they hold it on primarily moral grounds: those who work hard should enjoy commensurate rewards as a simple matter of fairness. Those who screw up should bear the consequences of their mistakes.

Economists have a similar principle, but they hold it for quite different reason. Economists believe that people respond to incentives, and so by tying a person's rewards as closely as possible to the social good or harm done, we induce socially optimal behavior.

In fact, economic logic often favors policies that violate conservative principles. A recent example were the bank bailouts in 2008. Many conservatives argued for allowing banks to fail on principle, but this would have been disastrous from an economic perspective.

Fundamentally, the principles of economics are about efficiency, not morality. This isn't to say that economists do not recognize moral justifications for public policy as legitimate.[2] It does mean that both political parties will find academic economists fickle friends if their platforms are fundamentally based on moral arguments, rather than arguments about how the world works. Economics is a science, not a moral philosophy.

[1] So before, were defendants presumed innocent by reason of insanity until proven sane? I suppose that is a question for a lawyer.

[2] An example of a case in which there is both a significant moral and scientific dimension is the death penalty. There is both a scientific and moral question here: Does the existence of the death penalty decrease crime? And does society have the right to execute its members? Someone could consistently believe that the death penalty reduces crime but it is immoral. Or the reverse.

Tucson Shooting

2011-01-11T09:41:00.001-05:00

I was going to make a post giving my thoughts on the Tucson shooting, but my views are perfectly summed up by this post by Nate Silver and this column by David Brooks.

My favorite quote from the Brooks piece:

If the evidence continues as it has, the obvious questions are these: How can we more aggressively treat mentally ill people who are becoming increasingly disruptive? How can we prevent them from getting guns? Do we need to make involuntary treatment easier for authorities to invoke?

These are the questions we should be asking. I do not like Sarah Palin, and I find much of modern political rhetoric abhorrent (and I agree that, while it comes from both sides, it is more common on the Right), but this incident seems to be completely unrelated to political rhetoric.

QE2: Should long-term rates go up or go down?

2010-12-15T16:55:00.003-05:00

Two weeks ago, I was explaining QE2 to my undergraduates. As I was doing so, it occurred to me that the logic of quantitative easing doesn't indicate whether interest rates will go up or down. Ironically enough, this issue has subsequently been raised by several commentators, who allege that the recent increase in long-term interest rates since QE2 began are an indication of the failure of the policy. Sadly, my students may be among them, since none of them asked about the issue.

In usual times, if the Fed wants to prop up the economy, it lowers the interest rate. Lower interest rates makes saving less attractive and borrowing cheaper, so it increases spending and investing. To lower interest rates, the Fed e-prints* money and purchases short-term treasuries. These purchases increase the demand for treasuries, which drives up their price. The price of an asset is the inverse of its return, so driving up the price is equivalent to driving down the interest rate.

The normal interest rate that the Fed targets is the interest rate at which banks lend to each other, known as the federal funds rate. When this rate hits zero, it may seem that the Fed is out of ammunition. However, there are many other interest rates in the economy, and there are spreads between these and the baseline federal funds rate. These spreads are determined by risk associated with these assets relative to the nearly 100% safe overnight rate. Long-term treasury bonds have two sources of risk: risk of default by the US government, and risk of inflation. So the interest rates on long-term bonds will generally be a bit higher than those on short-term bonds.

So when "normal" monetary policy has run out of ammo, the Fed can do "quantitative easing." Instead of e-printing money and buying short-run bonds, the Fed e-prints money and buys longer-term bonds. This increases demand for these treasuries, which drives up their price, which drives down their return.

So after a round of quantitative easing, we should see long-term interest rates going down, right? Not quite. You see, as I explained to my undergraduates, the purpose of QE is to avoid a deflationary scenario ala Japan. The Fed is trying to boost inflationary expectations. Inflation in the current environment is good for the same reason that deflation is bad: A good dose of inflation will boost demand and reduce aggregate real debt burdens. So interest rates will go up as inflationary expectations kick in, at least if QE is working.

Sadly, none of my students asked at this point, "But Mr. TA, if this all happens because interest rates are going down, then how are they also simultaneously going up?" No commentators asked Bernanke that either, which is why they're confused now that interest rates are going up.

What's going on here? Well, the nominal interest rate is the real return on the asset plus the expectation of inflation. So i = r + pi (where pi is inflation, or in this case expected inflation). The Fed's increase in the demand for long-term bonds has decreased equilibrium r, but it has also increased pi, so in principle i (the headline nominal interest rate) could go up or down.

But why can that happen at all? Doesn't the whole mechanism work by lowering the interest rate, and so if interest rates in fact increase doesn't that mean it isn't working? If the interest rate is where it started out, what has changed?

A lot. The interest rate may be the same, but the Fed has purchasing a quantity of bonds with money that it just e-printed. That money has crowded out other purchases, and those purchases have gone into other things. Hopefully things like business investment and consumer spending. If all the money had stayed in bonds, we would have seen interest rates drop by a lot. Instead we see interest rates rising, and because all that money is now in other things, it's increasing the odds of recovery and of inflation.

===================================
* Bernanke has been very adamant that the Fed is not "printing" money. They are merely electronically crediting the accounts of banks with the Federal reserve. Thus cash in circulation is not increasing, but only bank reserves. Of course, they're all part of the monetary base, but out of respect for the Fed Chairman, I'll refer to this as "e-printing." Hopefully that will satisfy Ben.

How Low are Savings Rates?

2010-11-07T22:03:00.004-05:00

One feature of the US economy over the past two decades has been its very low savings rate. This has led some to decry American profligacy and consumer culture.

While culture can certainly play an important role in economic decisions, other explanations have been offered. For instance, one common explanation is that China and other developing countries have been buying up lots of dollar assets in order to devalue their currencies, or as a safeguard against currency crises. This global "savings glut" drove down interest rates, leading to low savings rates in the US.

In this post I want to suggest an alternative explanation. Perhaps part of the problem is that our definition of "savings" is not broad enough.

We often think of savings as accumulating assets to finance future consumption. The motive could be to smooth consumption, or to take advantage of perceived high returns, or to build up precautionary savings as a buffer against uninsurable risk. However, it's possible that our focus on looking at the assets has blinded us to a lot of saving that goes on.

When we talk about "saving", the essential activity is the sacrifice of present consumption in favor of accumulating some asset to finance future consumption. Traditionally, we think of tangible assets (typically financial instruments), and thus we can approximate the value of consumption foregone by looking at the purchase of tangible assets by households.

However, there has been growing recognition among economists of the increasingly important role played by intangible assets in our economy. If intangible assets are important, then we are potentially missing a lot of household savings.

What sort of intangible assets do I have in mind? Well, what sorts of activities do people do to invest in their own capabilities (and thus value), and in the process forgo consumption today?

(1) Formal education -- people choose to invest time, energy, money, and foregone wages in pursuing an education.

(2) Unpaid internships -- Whenever somebody accepts a low-pay or no-pay job, internship, or volunteer opportunity, they are doing so because they're investing in some sort of intangible asset, such as experience or networks.

(3) Time spent acquiring or practicing skills informally on one's own.

(4) Participation in clubs, community events, or service projects to build your resume or network.

(5) Free or underpaid talks, writing, research, etc. to increase name recognition, prestige, and demonstrate skills.

(6) Blogging (sometimes)

These activities all represent investments in intangible assets such as skills, experience, knowledge, or relationships that one expects to have value. They represent foregone consumption, either of material goods or of time, and are a substitute for more traditional tangible assets. In other words, these activities are best thought of as purchases of another category of assets.

For example, although many graduates of elite universities have significantly negative tangible net worth due to student loans, we consider them quite wealthy because of their substantial stock of intangible assets. The value of an education at an elite university is not simply the knowledge and skills you acquire, but also the prestige, contacts, and benefits of being an alumnus (i.e. various University resources) that come with it.

I don't know what the US savings rate is when we factor in the value of foregone consumption and leisure in building intangible assets. It might still be lower than in the past, or it might be the same or higher. Maybe we still need to think about China's currency policies and our consumer culture, but I'm pretty sure we're missing a big piece of the story by ignoring intangible assets. The headline savings numbers are leaving out a lot.

Theory of the Nation

2010-11-04T22:43:00.003-04:00

Growth economists think a lot about questions like these:

(1) Why are there income differences across countries?
(2) Why do countries become richer (or poorer) over time?
(3) Is growth sustainable?

One question that growth economists do not generally consider is the sources of income differences within countries. This bothers me a bit, because I think that it implies the very strong assumption that the nation is the relevant level for thinking about economic growth. This assumption requires a "theory of the nation" to be justified.

My use of the phrase "theory of the nation" is a reference to what in economics is called the "theory of the firm." This refers to theories about what makes "the firm" a relevant level of economic analysis. This isn't immediately obvious. A single "firm" can include completely separate production facilities with little overlap. People used to just sweep this issue under the rug by using assumptions that made the nature of the firm irrelevant. More recently people have become very interested in this, because you need a theory of the firm if you want to think about things like in-house vs. out-sourced production, or the decision to partner with or purchase a firm overseas to produce inputs for your products, rather than just buying from them directly.

Likewise, I think that growth economists need to think more about our "theory the nation." What makes the nation the appropriate level for thinking about economic growth? There are a few arguments I can think of:

(1) Countries generally have similar legal and regulatory regimes, which may be quite different from other countries, but quite uniform throughout the country.

(2) There may be a distinct "national culture" that's important for understanding growth. (there's a parallel here with people who think that "corporate culture" is an important fact about firms)

(3) There's a lot more labor-market mobility within a country than between countries, because most people have social networks within their country but not in other countries, and language and cultural barriers may make them unwilling to move across national boundaries.

(4) Nations often have various barriers to trade - whether economic or geographic - that make them isolated to varying degrees.

These are all potentially good reasons for thinking about growth at a national level. However, I'm not fully convinced. My intuition is that there are probably a lot of similarities between the sources of wealth differences within a country and the sources of wealth differences between countries or over time. This may be particularly true for large countries, where different regions may have levels of cultural or geographic isolation comparable to those between some nations.

Health Care Reform

2010-04-18T21:10:00.004-04:00

Now that health care reform has passed, I thought I would make a post about what's in the bill, and why it's there.

The US health care system has two major problems. The first problem is a cost problem: health care costs are high and growing rapidly. There are many views on why this is the case. Some people believe that the insurance policies socialize costs leading to over-consumption. Others place the blame on lack of preventative care, high obesity rates, high malpractice rates, or malfeasance on the part of large companies, be they for-profit medical providers, insurance companies, or pharmaceuticals. Finally, a few economists have argued that health care is becoming more expensive because it's higher quality and we want more of it as we get wealthier. In other words, high costs are not a "problem" that needs to be fixed, they're simply the result of the market for health care doing it's thing.

The second problem is a coverage problem. There are many people who do not have insurance some of the time, and there are a substantial number of people who do not have insurance for long periods of time. Moreover, these latter people are among those who need care the most.

Health care reform was not about the problem of cost. Although there was much talk about costs, and the need to keep them under control, that was not the focus of the legislation. The cost-cutting measures included in the bill were purely cosmetic, intended to make the bill "deficit-neutral" for political reasons. The long-term problem of medical costs, and the related problems of an aging population and projected high costs of medicare, were not the subjects of this bill. Those will have to be dealt with at some point, but apparently now was not the time. This bill was about coverage.

In the US, individuals have health insurance coverage from three sources. A portion of the population is covered by the government (the elderly and disabled through medicare, the poor through medicaid). Much of the rest of the population is covered by employer-provided plans. The remainder must purchase insurance on the individual market.

The problems with the individual market are well-documented. Affordable health plans are available to purchase if you are young, healthy, and male. If you are not all three of these things, then health care is expensive, often prohibitively so. If you have a preexisting condition, your chances of being able to purchase coverage are essentially zero. Moreover, if you develop a serious medical condition while on an individual plan, your insurance company has a strong incentive to try to deny you coverage of expensive treatments, or to drop you entirely if they can find a way to legally do so. Thus health insurance is unavailable to those who need it most, and is often taken away from those who find they need it.

Given this situation, some sort of reform was necessary. At the beginning of the debate, some people suggested a fundamental reform in how we pay for medical care in this country. However, there were binding political constraints at work here. Rightly or wrongly, Obama felt that he couldn't touch people who were already generally satisfied with their coverage. That meant that anybody with insurance from the government or from their jobs were off the table.

There are basically three reasons people do not have medical insurance. Some people do not have it because they are young, healthy, and male. Although they can purchase good health insurance at affordable prices, they have no incentive to do so because they are unlikely to use it. Others do not have insurance because they are too poor to purchase it, although they would like to have it and are healthy enough to qualify for a reasonably-priced plan. Finally, some people do not qualify for any reasonably-priced plans because they possess severe (expensive) medical conditions.

The bill uses a hodgepodge of reforms to cover the uninsured. It tries to expand employer-provided health care by requiring more companies to offer insurance plans to their employees. It expands medicare and medicaid to cover more individuals, particularly more lower-income individuals. It an insurance mandate, requiring those who would like to not buy insurance to do so. It offers federal assistance to those unable to afford insurance so that they can afford to purchase insurance privately. Finally, regulations are put in place so that insurance companies can no longer deny coverage on the basis of preexisting medical conditions.

So what's the final analysis? Despite the rhetoric surrounding the bill, it really just amounts to extending the current system in a minimal way to cover the uninsured. One outstanding question is whether these regulations will end up destroying the individual market, or whether enough healthy individuals will enroll that the market remains profitable. There are serious adverse selection problems here, and given the forced blindness to riskiness of those seeking insurance, this is not a small concern. A key question is whether the individual mandate has enough teeth that people who don't want to actually sign up for insurance, which obviously will not happen if the mandate is found unconstitutional. I can envision a scenario in which it no longer is profitable to sell insurance to individuals at all, and insurance companies withdraw from this market entirely.

In the end, I think that this reform is better than nothing, since it should hopefully remedy some of the more extreme abuses of the individual insurance market, and extend coverage to some lacking it. However, it does not tackle the essential problems with the health insurance market or the problems arising from rising health costs and an aging population. The nation faces serious long-term fiscal challenges, and this reform does little to address them.

Irrationality and Framing

2010-03-27T11:10:00.003-04:00

Economists get a lot of criticism for assuming rationality. But what exactly does this assumption mean at the micro level?

On the level of consumer decisions under certainty, the key to "rationality" is transitivity. If someone chooses A over B, B over C, and C over A, then their decision-making process is not rational, in that we can't represent it as maximizing a utility function. If u(A) > u(B) > u(C) > u(A), then u(A) > u(A), which is not true of any real number.

Human decision-making is not transitive, however. Psychologists have conducted experiments in which people were asked to choose between pairs of goods with various characteristics. If you ask for enough comparisons, people will invariably slip up somewhere and choose A over B, B over C, and C over A.

I say "slip up" because when you bring such a decision to someone's attention, they won't try to defend it. Most people agree that transitivity is a minimum standard for "good" decision-making, and when confronted with intransitivity in their choices, they will usually reassess those choices.

So intransitivity clearly exists in decision-making. How would we try to model it? Suppose a consumer walks into a store with three products, A, B, and C. Suppose further than the consumer prefers A > B, B > C, and C > A. What will the consumer buy?

We can't answer that question without making assumptions about the buying process. For instance, suppose people decide what to buy using the following procedure:

Current_best = 0 (or some minimum standard)
For all products:
If product > current_best then current_best = product
Buy current_best

Using this procedure, people compare products sequentially. If they see A first, then B, then C, they will compare A to B (choose A), compare A to C (choose C), and buy C. If they see A C B, they will compare A to C (choose C), compare C to B (choose B), and buy B.

Thus the outcome depends on the sequence in which the goods are presented. In other words, framing matters.

Now any good psychologist or sociologist will agree that framing matters in perception, and thus in decision-making. But it's pretty difficult to model. If people behaved rationally, we could derive consumption as a function of wealth, prices, and goods available. If framing matters, we have to specify a decision-making framework, and then try to gauge the manner in which the choices are presented. That's a lot more complicated.

Now let's turn to a slightly different question. Suppose we now allow uncertainty. Expected utility theory makes the claim that we can model people's decisions under uncertainty as though they had utility functions defined over outcomes, and their utility function is linear in probabilities. This means that if the utility of outcome A is 100, the utility of a 50% chance of outcome A is 50.

This is a pretty extreme assumption, and is a lot more controversial among economists than rationality of consumer preferences. It's true that intransitivity can show up in certain experimental contexts, but these experiments generally result from forcing people to make lots of comparisons of things that they don't have strong preferences over (close to indifference), and confuse them enough that they make a mistake. By contrast, exceptions to expected utility theory not only occur in more contexts, but people are more apt to defend them when they are brought to their attention.

Let's give an example. Suppose that you are given a choice between playing lottery A and lottery B. In lottery A, you have a 98% chance of winning $10,000, and a 2% chance of winning nothing. Lottery B gives you a 100% chance of winning $9,000.

Most people will pick lottery B here. Although they give up $1,000 2% of the time, they don't want to risk winning nothing.

Now consider a choice between lottery A' and B'. Lottery A' gives a 49% chance of winning $10,000 and a 51% chance of winning nothing, while Lottery B' gives a 50% chance of winning $9,000 and a 50% chance of winning nothing. Which do you choose?

Most people will pick lottery A' here.

Now suppose someone offers the following game. He'll flip a coin. If it comes up heads (50% chance) you get to choose between lottery A or B, and if it comes up tails you get nothing. The only trick is, whether you choose lottery A or B before he flips the coin.

If you think about it, this game has the same payoffs as A' and B'. Yet most people make a different choice between A and B based on nothing more than whether there's a coin flip beforehand or not (or whether they're told about the coin flip).

The psychology behind this is quite interesting, but there is a clear parallel to rationality in consumer choice. In both cases, the effect is that people's decisions depend on how the problem is posed.

So are rationality in consumer choice and expected utility good assumptions? I think that they are reasonable assumptions in most economic contexts, because most in most economic theories only require rationality at the margin. However, irrationality does matter in some contexts, and in those context we have to find ways to model it. This requires understanding framing and its effects on decision-making.

When is Rationality a Good Assumption?

2010-03-21T10:12:00.006-04:00

Economists receive a lot of criticism for assuming that all agents in their models are rational. Such criticism is certainly not limited to non-economists; plenty of economists have expressed reservations about the extreme assumptions present in some economic models.

Rationality is convenient because it allows us to model decisions in a consistent manner without making extreme assumptions about short-sighted behavior. Pure myopia and linear decision-making is clearly incorrect, and hyper-rationality is a tractable alternative. So when is rationality a good assumption, and when is it a bad assumption?

Suppose agents' payoffs are a function of their own actions and also the actions of others. In economist speak, Ui = e(ai,a-i), i.e. utility of person i is a function of their actions (ai) and the actions of others (a-i).[1]

So given the actions of others, we can define a best response by person i. Let's call this function ai = BR(a-i).[2] Then the Nash Equilibrium is a* = BR(a*). In fact, given a few reasonable assumptions we can guarantee the existence of a Nash Equilibrium.

Now suppose that some agents don't behave as though they're maximizing the function U. Suppose they behave in a random manner. Will this significantly change the outcome?

Intuitively, this depends on the sign of the partial derivative of e with respect to both arguments, i.e. e12. If e12 > 0, then when others irrationally increase a, the rational agent has an incentive to increase a as well, or to "imitate" their irrationality. If e12 < 0, then the rational agent has an incentive to compensate for others' irrationality.

For a case in which e12 < 0, consider traffic congestion. Suppose that there are two roads going between two locations, route A and route B. Suppose that some drivers are non-optimizing and pick a route at random. Suppose that a few drivers are optimizing and listen to the radio to determine the route with the least traffic. When more irrational drivers pick route A, rational drivers take route B until travel time is equalized between the two routes. Thus even with a lot of irrational drivers, the rational drivers can restore equilibrium. This is an example in which a little bit of individual rationality can buy you a lot of aggregate rationality -- the outcome is the same as if everyone were rational and listened to the radio every morning.

For an example of the opposite, suppose that there are a number of firms that are deciding how to set their prices. Some firms are more productive while others are less productive, but they are all competing against one another. Firms want to set prices that maximize their profits, but their sales depend on how much higher their prices are compared to their competitors.

Now suppose there is a productivity shock such that all firms are less productive. Since it is now more costly to produce a given level of output, firms will want to increase prices. However, suppose some firms are irrational and don't change their prices. If the rational firms increase their prices as much as they would have in a rational equilibrium, they would lose too many customers to the irrational firms that kept their prices low. Therefore the rational firms do not raise their prices as much.

In the first case, you benefit by doing the opposite of others (e12 < 0). i.e. if other drivers choose route A, you want to choose B. Therefore we only needed rationality at the margin, since marginal rational agents act to arbitrage the market.

In the second case, you are harmed if you do something too different from others (e12 > 0), so rational agents have an incentive to imitate the irrational agents to a degree. In this case, you only need a little bit of irrationality to get a highly suboptimal outcome.

So in general, if you want to know whether rationality is a bad assumption, you need to determine whether rational agents benefit by doing the opposite of irrational agents, or whether they're better off playing along. In the former case, you're okay assuming rationality. In the latter, you have to find a way to model the irrational agents.

[1] I'm assuming that everyone has the same utility function for ease of exposition.
[2] BR is defined implicitly as that value of ai such that e1(ai,a-i) = 0, the first order condition for optimality.

What Economists Know About Recessions

2010-03-15T10:44:00.004-04:00

Robert Shiller has a new opinion piece out in which he says that economists didn't predict the current recession, don't know what caused it, and therefore can't know what the recovery will look like. He even compares our state of knowledge of macroeconomics to the Great Depression

Is this really an accurate picture of economists' understanding? What do economists know about the recessions?

In recent decades, economists have primarily used the "impulse-response" framework in thinking about business cycles. The economy is assumed to be subject to random shocks (impulses), which then spread and deepen into recessions through various propagation mechanisms.

The shocks can be of many kinds -- policy shocks, price shocks, demand shocks, technology shocks, and so forth. The precise nature of the shock is largely secondary; in this framework, the emphasis is on the propagation mechanism, not the shock.

As a result, the question "What caused the current crisis?" is simply not a question that economic theory is equipped to answer. If pressed, most economists would probably say something like this: "The current crises was caused when the housing bubble popped, precipitating a financial crisis and a severe recession." But economic theory can says little about the origin of the bubble, or the reason for the vulnerability of the financial system to housing.

It's likewise unfair to ask economists why they failed to predict the crisis. For one thing, most economists did recognize that the housing bubble posed a threat to the economy, and many predicted a mild recession. Few foresaw a deep recession simply because few had any idea of the vulnerability of the financial system to housing prices. Such vulnerability was simply not something economists were looking at.

All of this does not imply, however, that economists have nothing to say about the recovery. Our theories are not designed to predict recessions and financial crises, but they do say a lot about how recessions proceed once underway and what can be done to fight them. That isn't to say that there's a simple prescription or a single prediction offered by economic theory -- here economists truly are of many different opinions. This is unsettled science, and we suffer not from having nothing to say, but from having too much.

A Taxonomy of Business Cycle Theories

2010-02-10T10:58:00.005-05:00

A lot of economists think that the government should be doing more to help the recovery. Of course, this is certainly not the unanimous view, but I think that we can definitely call it the majority view.

Although the public debate is highly politicized, the disagreement among economists is not at root political. Rather, it reflects a long-running disagreement among economists about the nature of business cycles.

Most economists agree that the current crisis was precipitated by the collapse of an asset bubble. At first the recession was mild, as only sectors immediately connected to housing suffered much. But then the crisis moved to the financial system, and the recession became a financial crisis, the most dangerous form of economic downturn.

Although the financial mess was (mostly) cleaned up, we didn't return to business as usual. Rather, we're suffering a deep and painful recession, with the labor market hit particularly hard. Although the onset of crisis went from housing to finance to the rest of the economy, curing housing and finance are clearly not enough. The world economy is suffering. How did this happen?

Here's where the narratives diverge. Most business cycle models take occasional large shocks as a given, and then explore mechanisms by which these shocks generate observed business cycle dynamics. Policy prescriptions when facing a major recession will depend on what you believe are the principal propagation mechanisms at work.

A lot of propagation mechanisms have been proposed. One is the Keynesian mechanism: demand. All the sectors of the economy are linked. In some cases, they are linked because the output of one sector is bought by another. In other cases, the link is simply consumer demand. When a sector contracts, it fires workers and decreases income, which leads to decreased demand for consumer products which affects the whole economy. If the economy isn't able to adjust efficiently, this leads to further layoffs and a further reduction of demand.

Old-fashioned Keynesian models relied on unrealistically myopic and simplistic behavior by economic agents to generate these dynamics. Once you assume that people aren't idiots who behave the same way regardless of the situation, it becomes much harder to generate the Keynesian story. However, many modern models have managed to preserve demand as a basic propagation mechanism, and thereby retain a distinctive Keynesian flavor. Many of these models emphasize nominal rigidities, information asymmetries, and limited rationality to generate propagation.

The chief alternative to demand-based propagation is to emphasize the real structural complexities of the economy. When a shock hits, there is a reallocation of economic resources. However, such reallocation is not instantaneous, because things take time in the economy. Production is not instantaneous, and labor and capital markets face various non-convexities and irreversibilities that generate non-linear dynamics. The result is an adjustment process whose length potentially increases with the size of the shock.

What you believe about recessions has major consequences for policy prescriptions. If you think that demand shortfalls are the main mechanism, then government needs to pick up the slack. Increasing government spending and hiring more government workers will increase aggregate demand, and bring the economy back from the brink.

However, if you think that what's going on is a reallocation process, then government spending could be counter-productive. Does that mean that there's nothing government can do? Not necessarily. As I explained in my previous post, the burden of debt represents a potential drag on growth. This goes equally well for reallocation, which is basically what growth is. So anything we can do to reduce the burden of debt could help us to recover faster. One option is converting private debt to public debt by deficit-financed tax cuts. If these tax cuts encourage hiring, so much the better.

It looks like president Obama is leaning more towards tax cuts for any further job bill. This is probably based on political rather than economic calculus, but it represents a victory for reallocation perspective nonetheless.

Of Debt and Stimulus

2010-02-08T13:38:00.003-05:00

I find the issue of debt fascinating, because debt is a human institution with essentially no physical representation. It's a fiction that we use to organize human affairs by promising goods and services now for goods and services later.

It seems strange that the debt situation could significantly impact growth. Afterall, growth is driven by real things like technology and machinery. Why should the partcular arrangement of paper IOUs in the economy make such a big difference? Yet there is a growing belief among economists that the international debt situation could lead to significantly slower growth over the next 5-10 years. See for instance this interview by Carmen Reinhart.

When an economy has a lot of bad debt, as ours does, it can significantly impede the coordination required for economic growth. Companies put their earnings into paying down debt, and forego investments in expanding production capacity or research. Individuals restrict consumption and investing in education in order to pay down debt. Banks curtail lending, tightening credit to businesses and consumers. Finally, declining tax revenue and crisis-related expenditures increase public debt, which creates uncertainty about future spending, taxes, and various forms of default.

So how do we deal with the debt? One option is to bail out affected institutions, cut taxes, and otherwise inject money into the system. This money can be used to clean up private balance sheets and get the private sector moving again. If the government borrows to finance these actions, this amounts to replacing private debt with public. If the government prints money to cover these actions, there will be inflation reducing the value of debt. This is equivalent to nationalizing the debt, and then defaulting on it.

Any attempt to nationalize the debt runs a very serious risk. If public debt becomes too bad, it can become an even larger burden on growth than private debt. Any effective default via inflation would set a bad precedent, and fear of such could lead to a run on the currency. Attempts to clamp down on speculation to prevent panicked investors from inducing a crisis could have substantial collateral damage to legitimate investors and the functioning of the financial system.

Ultimately, we must move debt to those institutions best able to bear it. Thus we must determine what institutions are best able to bear the debt, how to transfer it to them, and what to do with it once it's there. And we must do this while remaining mindful of the far-reaching implications of our actions to incentives and expectations.

Poverty (Part 1)

2010-01-09T19:16:00.014-05:00

Why is there poverty?

This wouldn't have been an interesting question 1,000 years ago, or even 300 years ago. Then the answer would have been obvious: agriculture was not productive enough to sustain high per capita wealth. Any advances in technology that increased productivity created only temporary wealth, since population increases would quickly consume the surplus. Good land was in fixed supply, and crop yields were barely high enough to feed the farmers. Thus the great majority of humanity was engaged in subsistence farming. To the extent that anyone had any wealth to speak of, it was because he (and it generally was a he) had extracted it from his fellow man through unjust social structures.

This description does not apply to the world today. Much of the world's population has escaped from the Malthusian trap. This has happened for two reasons. First, the rate of technological advances has increased significantly. Second, population has ceased to grow in proportion to wealth. In fact, population would be falling in most wealthier parts of the world but for immigration.

Yet poverty still exists. There are many countries that still experience widespread poverty, and there are still segments of the population of wealthy nations that live in poverty. Why is this?

As I've mentioned before, persistent poverty shouldn't exist according to standard economic theory. It is indeed a big mystery. Why don't poor countries simply adopt the production technologies of the first world, and quickly achieve comparable levels of wealth? Why don't poor individuals simply imitate the rich? Sure they would be starting with less wealth, but why shouldn't be able to achieve comparable incomes?

One might argue that you need wealth to get wealth. You need enough money up front in order to build a factory of get a graduate degree. The poor are simply unable to pay the start-up costs necessary to be wealthy. But then why can't individuals borrow against future incomes to undertake the necessary investment? One might argue that credit markets are incomplete and such borrowing is not possible, but why is this? There are massive gains possible should the poor simply adopt the same production and investment strategies as the rich, so why hasn't the market provided mechanisms for such investment? Why hasn't the international community made this a huge priority?

Unfortunately, they have and it hasn't worked. It turns out that ending poverty is not as simple as the poor simply acquiring the same skills and production technologies as the rich. Or rather, it turns out that the poor acquiring those skills and technologies is a very difficult proposition.

The is because technologies and skills are highly contextualized. One cannot simply uproot a computer and ship it to a third world country, or even to a poor family in America, and expect it to be as productive as it was before (although most growth models say it should be more productive there). The market put that computer where it was for a reason. The computer is the product of a complicated system. Its creation involved researchers who invented the concepts of computation in general, programmers who created the software for the needs of the company or individual that owned it, companies that designed and manufactured the silicon chips and other parts of the hardware. That particular computer was designed to be used by highly trained individuals in the precise context in which it was found.

It turns out to be deeply significant that the wealth of the world is produced. Technology is not a generic residual, and capital is not an undifferentiated mass experiencing diminishing marginal returns. Rather, technology is process and product innovation, used to create particular products which are demanded by particular individuals and companies. While a good produced in the US can be sold in Africa, the production process that produced the good cannot be so easily shipped overseas.

So far I've merely used a lot of words to say, "It's complicated." If I were a Libertarian, I would conclude by singing a song about the virtues of the free market, and tell everyone to leave it alone to do its magic. But that would be folly. The fact that science is difficult is not an excuse for not doing it. The problem of poverty is not as simple as it at first appears, but that does not mean that we should conclude that it is unsolvable, and consign the economist to merely describing its complexities and extolling its wonders. Poverty in the midst of plenty is evidence that parts of our economic system are deeply broken, and we ought to try to fix them.

Housing and CRE

2010-01-08T12:29:00.005-05:00

Paul Krugman argues that the existence of a bubble in commercial real estate (CRE) proves that the housing bubble was not driven by subprime mortgages, the community reinvestment act, or predatory lending. His argument is that these would drive up housing prices, but not CRE.

I do not find this argument convincing. Prices of commercial and residential real estate are related on the supply side. A bubble in residential housing could drive up the cost of commercial real estate by driving up construction costs and bidding up the prices of properties.

As the graph above indicates, movements in CRE prices have lagged housing prices over the past decade. This suggests to me that the CRE bubble was a consequence of the housing bubble, not jointly produced by the same underlying cause.

Consumption Behavior

2009-10-20T16:20:00.004-04:00

How do people decide how much of their income to spend and how much to save?

This is a pretty fundamental question in Economics, and has been the subject of a lot of theoretical and empirical literature over the past several decades. The starting point for much of modern theory is Milton Friedman's permanent income hypothesis, which concludes that people should consume at their "permanent income," adjusting consumption only in the face of unanticipated shocks to lifetime income, and smoothing over predictable changes. In other words, people should save money when they are earning above their long-run "permanent income", and borrow when they are earning less.

However, the empirical evidence does not back this up. While people do smooth over some anticipated changes in consumption, they do not do so to the extent implied by theory, and in many cases fail to smooth at all.

Economists have basically taken two approaches to resolving this problem. The first is to attempt to model human irrationality. This is a difficult task because irrationality is such a broad term. "Rational" behavior is pretty easy to define since it represents a definable subset of behavior-space. "Irrational", on the other hand, is very broad, since it's defined as all the rest. I'd put models of bounded rationality in this category, along with models with loss aversion, models with habit-formation, and so forth.

The second approach is to posit failures in the economic system that prevents people from smoothing consumption as much as they would like. This includes liquidity constraints and differential rates of borrowing and lending.

Somehow I don't find myself very comfortable with any of these approaches. Somehow, this looks an awful lot like adding epicycles to epicycles. I feel like there should be an elegant model of consumption that can readily describe observed behavior without needing to append all sorts of additions and special cases. Maybe our failure to cleanly match the data suggest a more fundamental flaw in our theory of economic behavior.

Or maybe I'm just falling for a reductionist bias towards elegant mathematical models. Maybe economic behavior in complex society is, well, complex. That's also possible.

2009-10-12T20:38:00.002-04:00

Elinor Ostrom and Oliver Williamson have won the Nobel prize in Economics. They're responsible for pioneering New Institutional Economics. Paul Krugman has a good summary of the significance of their work.

The papers cited by the Nobel committee in their decision were published between 1971 and 1990. This is typical of the Nobel prize in economics, which is typically awarded for work done many years before. This is done because the ultimate significance of a body of work is not always readily apparent. Often promising and much-cited papers (or application essays) end up not being as meaningful in retrospect.

Apparently, the Nobel peace price is awarded according to a much different standard. In explaining the decision to give the award to President Obama, Thorbjorn Jagland, the Nobel committee chairman, said:

"The question we have to ask is who has done the most in the previous year to enhance peace in the world... And who has done more than Barack Obama?"

If this is indeed the criterion by which the Nobel peace prize is awarded, it is very different from the other prizes. Personally, I disagree with this philosophy. I would prefer that the peace prize be awarded to people for work done many years before.

This would serve two purposes. First, it would prevent current politicians from receiving the award for work on current issues. Granting the award to present politicians gives appearance that the Nobel committee is using its power to make statements about controversial political issues. Secondly, it would avoid the embarrassment of granting the award to politicians for work that ends up being short-lived, or who go on to make a travesty of the award in later years.

I believe that the way the peace prize is awarded at present gives a strong impression that the Nobel committee is using the prestige of the award to make political statements. Further, the spotty history of the award shows that the second concern is not idle speculation. Altogether, I have to conclude that the peace prize cheapens the institution, which is unfortunate because I think having it is a great idea.

Sleep Schedules

2009-10-08T21:02:00.002-04:00

Noon is the middle of the day, and midnight is the middle of the night. One would expect diurnal animals to keep symmetric hours. If we assume that 8 hours of sleep each night is optimal, that would mean going to bed at 8pm, and getting up at 4am.

Yet this sleep pattern is very unusual. Most people would probably consider anything before 7am "early." Yet if one rises at 7am, one has already missed an hour of sunlight at the equinox.

So why do we keep such an unnatural sleep schedule? And does it have any adverse effects on our health?

John Taylor on the Minimum Wage and Unemployment

2009-10-03T21:20:00.005-04:00

John Taylor points to an editorial in tomorrow's WSJ concerning the recent minimum wage hike and high teen unemployment. He also points to a discussion of empirical research on the minimum wage from his introductory economics textbook. He mentions that the research generally shows a significant impact:

But as is so often the case in economics, not all economists agree, so the box also describes some contrary findings by David Card and Alan Krueger. I side with Neumark and Wascher in this debate, but you can read the editorial, look at the diagram, read the box, and draw your own conclusions.

John Taylor is the originator of the so-called Taylor Rule, that closely predicts actual Fed policy. Papers by Neumark and Wascher are here and here.

2009-10-03T21:20:00.001-04:00

Why Do Lawyers Run the World?

2009-10-01T20:35:00.003-04:00

The governments of the world are run by politicians. In the first world, most politicians are lawyers by training.

Why is this so? Why aren't countries run by experts in policy or public administration? It's because our politicians are elected by popular vote. The ideal training for a democratic system is in persuasion of the median voter.

In public policy schools, they teach you how to make good policy. In economics departments, they teach you how the economy works. In schools on public health or public administration, you learn how various parts of the system functions. These degrees teach you to choose correct policies, and to think in such a way as to persuade experts in those fields. This is all very useful if you're running a country, but it doesn't make you particularly good at persuading the average voter that you would be good at running a country.

Lawyers are taught to argue cases, and they're taught to argue cases to non-technical audiences. It's true that not all lawyers are trial lawyers, but the field is still predicated on making a case, not finding the right case to make. Those are the skills they teach you in law school, and those are among the skills most useful to being a politician. They are not particularly helpful to being a lawmaker.

What are the consequences of this state of affairs? At one time I agreed with Winston Churchill's assessment that democracy is the worst possible system of government except for all the others, but now I'm a lot less sure. There probably is a viable alternative that would be a significant improvement, we just haven't thought of it yet.

Isn't It an Amazing Coincidence...

2009-09-29T21:58:00.003-04:00

...that those who believe the benefits of a policy are high, tend to believe that its costs are low?

A few examples prompted this thought, but similar coincidences are quite common among social commentators, along with the converse coincidence that policies with limited benefits (invariably advocated by the "other side") have high costs.

I sometimes feel quite lonely considering questions without a clear winner.

Health Care -- What's Next?

2009-09-24T17:33:00.004-04:00

I think that some form of the Obama plan will pass soon. That means we will have everyone covered, with regulations preventing insurance companies from rejecting applications or refusing to cover treatment due to pre-existing conditions, and subsidies so that everyone can afford insurance. We may or may not have a public option competing with the private plans.

What happens then?

If recent research is correct, we'll continue to have poor health and rising costs. This is America's poor health is principally due to lifestyle choices, not health insurance coverage. Simply put, Americans are fatter and work more than in other developed countries.

The Obama plan does not address these issues. Moreover, it's unlikely to lower costs. It's true that a public plan may cut into insurance companies' profit margins, but high profit margins are responsible for very little of the high costs of health care. The plan's regulations about pre-existing conditions will likely increase insurance premiums, since companies will no longer be able to refuse to take on expensive customers. If the public option is not subsidized by public funds, then it will not be control to control these rising premiums. If the public option has access to public funds, then the costs will still exist, but will be passed on to the taxpayer.

What will happen when our health stays the same or worsens, even as health care costs rise? Will the Obama plan be considered a failure? Will anyone be able to explain why it failed, and offer an alternative?

In the current market, most people get their insurance through their jobs, and the costs are the same for everyone. Thus nobody has a financial incentive to improve their health, since they know that others will pay for their treatment. If we're serious about improving health and lowering costs, we need to stop separating people from health care costs.

This should be done by establishing a better functioning market. Insurance companies have a valuable role to play in pricing risk and encouraging healthy behavior. If insurance companies are able to charge premiums based on measures of health, such as BMI, blood pressure, and cholesterol, people would have a financial incentive to control these factors.

If the cause of our poor health is our poor decisions, not limited coverage, then we need to make sure that people bear the full consequences of their decisions. That's how markets work to encourage efficient behavior, when all costs are fully internalized by decision-makers.

Research Interests: A Few Themes

2009-09-21T19:17:00.008-04:00

I've been thinking a lot lately about my research interests. I've already discussed my Big Question, but that's more of a life-long calling rather than a PhD dissertation.

To help organize my thoughts, I've been scribbling down ideas as they enter my head. In this process, a few themes have emerged. I thought it would be helpful to write them down here.

1. Heterogeneity
I don't think economists have done a good job dealing with heterogeneity in economic models. This is partly a failure of data, and partly a failure of theory. I love theory, but I'd like to work on both. To paraphrase James, "Theory without empirics is dead."

2. Rationality
I'm really interested in how people really make decisions, and how that affects the economy, all the way up to the macro level.

3. Economic Growth
Very broad, but very important. I have some ideas on how to improve endogenous growth models. Hopefully these ideas are useful and new, not tried and failed.

4. Inequality
I have a lot of ideas about how inequality can persist even if markets are generally functioning well.

5. Information
I'm pretty interested in how information is created and propagates through the economy.

6. Equilibria
I have an intuition that the equilibrium paradigm in economics has created a blind spot. I have some ideas about how to approach this.

Teenage Unemployment -- a followup

2009-09-13T00:10:00.010-04:00

Nick comments on my previous post about teenage unemployment:

I also wonder how many college graduates and otherwise-out-of-work people are occupying jobs usually taken by teenagers.

Intuitively, it might make sense that teenagers are hit harder during downturns since they experience higher job turnover. To distinguish between those, we need to look at how teenage unemployment compares to the overall rate during past recessions. Let's look at the data:

(Source)

My observations:
(1) Teen unemployment is generally higher than overall unemployment.
(2) Teen unemployment tracks the business cycle, and experiences larger swings than the overall unemployment rate.
(3) The teen unemployment rate seems to be generally increasing from 1948 until the early 1980s, at which point the trend stabilizes.
(4) Teenage unemployment is at its highest level on record, although overall unemployment is a little below its highest level in the post-war period.
(5) The 1992 recession also seems to have hit teenagers especially hard.

Now let's look at the the federal minimum wage looks like:

(Source)

The general features:
(1) The minimum wage increased faster than the rate of inflation from 1948 until around 1980.
(2) After 1980, the minimum wage was raised only a few times in the 1990s, and well short of inflation, and so deteriorated in real terms during this period.
(3) The minimum wage has been increased the last few years.
(4) The 1992 recession, which hit teenagers particularly hard relative to the overall workforce, was also preceded by a minimum wage hike in 1990.

I find it interesting that the overall trend in teenage unemployment seems to track the overall trends in the minimum wage. I would say that the minimum-wage-hikes-raise-teenage-unemployment hypothesis holds up well under a cursory examination of the data.

Teenage Unemployment

2009-09-10T17:18:00.010-04:00

One of the most basic concepts in economics is supply and demand. The idea is illustrated in the graph below:

The demand curve represents the amount of a good purchased at a given price, while the supply curve represents the amount of a good produced at a given price. When prices increase, suppliers sell more and consumers buy less, so the demand curve slopes downward and the supply curve slopes upward. The intersection of the supply and demand curves is a unique equilibrium price P*, at which the amount supplied and the amount demanded are exactly the same. The equilibrium quantity is represented by Q*.

Suppose that government passed a law setting the price to some level above P*. The supply and demand model predicts that this will result in less demand and more supply compared to the equilibrium. Since the total amount bought can't be more than the least supplied or demanded, this results in a shortfall, in this case a shortfall in demand. This situation is illustrated by this chart:

Here, P** is the government-mandated price, and Q** is the quantity of the good supplied at P**. Since you can't sell more than is being bought, this shortfall in demand results in a lower quantity of goods bought in the market than in the absence of the price control.

Note that if the government had set a price lower than the equilibrium price, there would have been a supply shortfall. This would still result in less of the good being bought, but in this case because less was produced at the lower price.

This simple prediction, that price controls leads to demand or supply shortfalls, has been borne out in the analysis of many markets. One particularly contentious application is to the labor market, since supply and demand suggest that minimum wage laws increase unemployment.

Over the past few years, there have been several increases in the minimum wage. Simultaneously, there has been a huge increase in unemployment. Of course, most of this unemployment is attributable to the recession, but perhaps some of it is due to the minimum wage. How might we determine this?

The labor market is not homogenous, and most workers command better than the minimum wage anyway. So who actually earns the minimum wage, and is likely to be affected by an increase? It turns out that many minimum wage workers are teenagers. Teenagers generally fill part-time cashier-type jobs, and the pay is commensurate to their skills and experience.

So how has teenage employment held up during the recession? Not well, it turns out. In fact, the teenage unemployment rate has hit 25.5%, compared to an overall rate of 9.7%.

Score one for Economic theory.

Health care Thought of the Day

2009-09-04T09:26:00.003-04:00

An oft-repeated critique American health care is that the US spends more per capita on health care, yet has a lower life expectancy than many other countries.

As I've said before, I'm in favor of some sort of health care reform on both efficiency and moral grounds, and when I first heard the above critique, I found it very compelling. However, recently a number of economists have been questioning its basic premise.

Essentially, these economists argue that:
(1) The US spends more on health care because we are richer than everyone else, and health care is a luxury good.
(2) The US health care system isn't that bad in terms of quality of care. In fact, if you look at medical outcomes from the same procedures, the US is among the best.
(3) Our low placement in international health system rankings is largely due to the number of uninsured in the country, not the quality of care provided to those who have insurance.
(4) Our relatively low life expectancy and other health indicators is secondary to our high obesity rate, not poor quality of care.

Followup to SAT Scores and Income

2009-08-30T10:51:00.003-04:00

Greg Mankiw has responded to criticism of his post on SAT scores and parental income. He points to a paper by Dartmouth's Bruce Sacerdote. The abstract reads:

I use a new data set of Korean-American adoptees who, as infants, were randomly assigned to families in the U.S. I examine the treatment effects from being assigned to a high income family, a high education family or a family with four or more children. I calculate the transmission of income, education and health characteristics from adoptive parents to adoptees. I then compare these coefficients of transmission to the analogous coefficients for biological children in the same families, and to children raised by their biological parents in other data sets. Having a college educated mother increases an adoptee's probability of graduating from college by 7 percentage points, but raises a biological child's probability of graduating from college by 26 percentage points. In contrast, transmission of drinking and smoking behavior from parents to children is as strong for adoptees as for non-adoptees. For height, obesity, and income, transmission coefficients are significantly higher for non-adoptees than for adoptees. In this sample, sibling gender composition does not appear to affect adoptee outcomes nor does the mix of adoptee siblings versus biological siblings.

I once again confess myself surprised at the size of the genetic effect. I would have guessed that whether your adopted mother graduated from college would be more important than whether your biological mother graduated from college. Instead, genes matter four times as much. Wow!

Family Income and SAT Scores

2009-08-29T16:46:00.023-04:00

Greg Mankiw points to a graph comparing SAT test scores by family income:

A lot can, and has, been said about this graph. Rather than comment on whether rich people are smarter than poorer people, I'll just point out the interesting shape of the graph.

The first thing I notice is that there seem to be decreasing marginal returns in test scores up to about $120,000/year. Throughout this region, the graph is not only increasing, but it appears to be "flattening out". In calculus terms, the second derivative is negative, and the graph concave down.

The second thing I notice is that there is a big jump in the very last income category. This suggests that the correlation between wealth and test scores doesn't flatten out shortly after $200,000. Rather, if we were to add several additional income categories, we would see average test scores continue to increase. This suggests that $200,000 is not a natural stopping point. I'm very curious what the shape of the graph is past $200,000. Does it eventually flatten out? If so, at what score and wealth level? Do the children of the super-rich averaging 590 on their math SATs? Or are they scoring 650, 700, or 750? We don't know.

In fact, eyeballing the graph with the last data point removed, it looks to be roughly linear from $110,000 to $175,000. It definitely looks like the increase from $130,000 to $150,000 is no less than the increase from $110,000 to $130,000.

The failure of the graph to flatten out suggests that even among the rich, higher wealth corresponds to significantly higher test scores. For instance, kids whose parents earn $120,000 - $140,000 average a 1060 on their SATs, while kids whose parents earn $140,000 - $160,000 average nearly 20 points higher.

Given this, I have difficulty seeing how the causality could run from wealth to test scores in this region. How does an extra $20,000 for a family already making over $120,000 increase SAT scores by 20 points? Is that small increase in wealth really resulting in significantly better tutoring, schools, and enrichment? It seems much more plausible to me that, for family incomes greater than $100,000, the correlation is driven by inherited ability.

Why Have Private Health Insurance?

2009-08-26T21:01:00.010-04:00

Several prominent economists have argued that a public option in health care would eventually lead to full government-provided health insurance, since for-profit private companies would not be able to compete with a public company enjoying explicit or implicit federal subsidies. For instance, here's Greg Mankiw:

An important question about any public provider of health insurance is whether it would have access to taxpayer funds. If not, the public plan would have to stand on its own financially, as private plans do, covering all expenses with premiums from those who signed up for it.... In practice, however, if a public option is available, it will probably enjoy taxpayer subsidies.... [Such] subsidies would prevent a public plan from providing honest competition for private suppliers of health insurance. Instead, the public plan would likely undercut private firms and get an undue share of the market.

I basically buy this argument, but what exactly is wrong with public health insurance?

To be clear, I'm not talking about publicly administered health care, in which hospitals are government-operated, and doctors are government employees. I'm just talking about public health insurance, in which the government uses premium payments, taxes, or public debt to pay for our health care costs from private providers. What's the advantage of private companies here?

Insurance plans basically compete on two dimensions: price and quality of coverage. In the case of health insurance, price includes the monthly premiums, plus deductibles, copays, etc, while coverage includes what the insurance plan will cover. In addition, companies compete on providing good service to customers, which basically covers how well the company treats you when you call.

Since there isn't a material good here, it's hard to argue that the government would produce a shoddy product. So what's the value added by the private sector? I think it basically comes down to two things: accurately pricing risk, and preventing fraud. Private companies have a profit incentive to do both well, while a public company would have a bureaucratic incentive to do neither.

First, pricing risk. In general, insurance companies would like high-risk people to pay more than low-risk people. This satisfies both a general sense of fairness as well as the more practical purpose of keeping people from passing their risk onto others. To take a concrete example, one might expect health insurance companies to charge higher premiums to obese customers or smokers, both of whom are at higher risk of developing a host of expensive medical conditions. This would encourage people to reduce obesity and smoking due to the medical costs, rather than passing those costs onto others.

Of course, there are problems with pricing risk. In the case of auto insurance, it's pretty reasonable to charge higher premiums to people who get in lots of accidents. In contrast, people aren't comfortable with charging higher premiums to people who develop expensive medical conditions, since these aren't generally considered to be your fault. Most people are okay with penalizing smoking, since it's viewed as within people's control and "sinful" anyway, but what about cancer patients? Isn't the whole point of insurance to cover expensive medical treatments after you develop a nasty ailment? Essentially, the question is how to distinguish between "risk factors" for which we want to charge higher premiums, and "pre-existing conditions" that we want to cover with no questions asked?

Most likely, a public company would cover everyone at a flat rate to avoid public outcry. There might be a few "sin taxes" on particularly egregious or anti-social behaviors like smoking, but others such as obesity probably wouldn't make it. By contrast, a private company would have a strong incentive to determine who is more likely to need expensive medical care, and set premiums accordingly.

The second service provided by private insurance companies is denying fraudulent claims. Insurance companies have a strong incentive to police claims carefully, since any payment of a claim is a loss. A public plan would have little incentive to catch fraudulent claims, because it has no profit incentive to avoid the costs. Moreover, any attempt to deny fraudulent claims runs the risk of denying legitimate claims as well. A public plan would have a high political incentive to avoid accidentally denying a legitimate claim, even if that means accepting many fraudulent claims. Private companies have a different set of incentives, leading them to deny many fraudulent claims, and also some legitimate claims.

So where does that leave us? There is legitimate value added by private insurance companies that a public company would have little incentive to provide. However, we already regulate some of these services, and if the current health care reform bill passes, we will regulate much more. For instance, the bill would prevent companies from charging higher premiums for pre-existing conditions, limiting the potential for risk pricing. It would likewise reduce or eliminate the possibility of denying claims, which would likely make fraud prevention more difficult.

Given that both of the services provided by insurance companies will be highly regulated, there may be little lost if we switch to publicly provided health care. However, I personally think that the accurate pricing of risk and denial of fraudulent claims are valuable services. The profit motive causes some negative results, but we shouldn't abandon the model. We just need to regulate those areas where there is fraud.

How should we go about regulating medical insurance? Based on the discussion above, the two areas of concern are pre-existing conditions and denying legitimate insurance claims. The first should be handled by:
(1) Mandating that everyone purchase health insurance.
(2) Mandating that the initial contract spell out the rules by which the premium may change in the future.
(3) End employer-provided health care -- people should be able to stick with the same plan from cradle to grave.

One might argue that (2) wouldn't necessarily end up penalizing things like obesity and smoking. After all, someone who was overweight or smoked would never sign up for a plan that they knew would charge them a higher premium. However, if some plans charged higher premiums on this ground, while others did not, people who were not overweight and were non-smokers would go to the plans that offered them lower premiums, forcing the other plans insuring riskier patients to charge higher premiums to avoid taking losses. So we would end up with those sort of premiums anyway. By contrast, nobody would sign a contract that they knew would refuse to cover them if they developed cancer or diabetes, or any other particular condition, so such a market would hopefully be able to distinguish between risk factors and medical conditions.

My proposal for fraud prevention is to shift the power to deny claims from the health insurance companies to the courts. That is, a health insurance company would be on the hook for any claim unless they brought civil charges against the claimant for filing a fraudulent claim. Needless to say, such a civil suit would pay significantly more than the value of the claim to make it worthwhile and to discourage fraud, and those filing fraudulent claims would face criminal prosecution as well.

I believe that such a health insurance market would be much more efficient than both the current system and those under consideration in Congress. Note that, although I have proposed more regulations and mandates in some areas, the overall effect is actually less government interference in general, in large part because of the elimination of the ridiculous subsidies for employer-provided health care. That's some cruft that it's high time we got rid of.

***************************************

I should note that I'm not addressing the question of incomplete medical coverage here. That's a separate issue, and as I've said before, I think it's high time we provided insurance for all uninsured Americans. I suppose that would be done via some sort of means-tested vouchers, unless we also adopted my tax proposal.

I'm also not talking about the host of other issues related to medical care, including quality of care, rising costs, an aging population, the pharmaceutical industry, obesity, and inequality in general. Those are other issues, quite separate from the insurance market. Although all economic questions are just variations on the same theme, when you come down to it.

Economics of Solar Panels: Dynamic Edition

2009-08-11T17:29:00.062-04:00

Scott Adams (the creator of Dilbert) has an interesting post on the economics of solar panels. Scott writes:

Is it economical to install a solar power system (photovoltaic) for a new home?

Assume you will live in this home for the next 30 years, you're in California where the sunlight is plentiful, energy costs are high, and the government is offering rebates. You run the numbers, and as long as the cost of the system is wrapped into your mortgage, you are saving cash from day one. Ta-da! It's good economics, right?

Not so fast. Economists consider all alternatives, and one of the alternatives is to wait a few years and then add solar power to your home, when the systems are likely to be far more efficient and much less expensive. If you wait, you run the risk of losing any government rebates, and there's an economic penalty for not wrapping the cost into your original mortgage. But waiting still makes sense if the new system is twice as efficient.

Scott does a good job of laying out the consumer problem, which is how much of a particular durable good to purchase at a given price when the consumer anticipates the quality-adjusted price to fall.

Let's extend the question a little by considering the firm's problem. The firm must decide how much money to invest in research that will reduce its costs in the future. It knows that if it reduces cost, it will increase profits, and thus it will increase investments to the level at which the discounted marginal increase in profits in the next period is equal to the marginal cost of additional investment.

However, the firm knows that reducing costs will also reduce the profit-maximizing price, since a firm sets price such that profit margins equals the inverse of the price elasticity of demand. Given the consumer problem outlined by Scott, reducing the price in the future will reduce demand now, resulting in lower profits in the near-term. Moreover, the firm can't make a credible commitment to charge a higher price in the next period, because the consumer know that once it gets to the next period, the firm will charge the profit-maximizing price. Thus the firm will invest less in R&D than if demand was unaffected by expectations of future prices (e.g. if consumers were credit-constrained).

The consumption of solar panels has positive externalities, so society as a whole would benefit if more solar panels were sold. Standard economic theory recommends that externalities be internalized through taxes and subsidies (negative taxes). Thus, the government should subsidize the production of solar panels. Moreover, since the firm anticipates lower costs in the second period due to its research, it will anticipate a smaller subsidy as well. This expectation acts as a tax on R&D, and will lead the firm to invest less.

The government might like to promise to set a higher subsidy in the next period in order to stimulate more investment. However, such a promise would not credible, since the socially optimal subsidy in the future is lower, not higher. Thus the government must be able to precommit to a higher subsidy by, for instance, passing a law setting subsidies for the next 20 years that cannot be modified.

Incidentally, this is a general principle that, in a dynamic problem, there are often benefits to a system in which government can make a credible commitment to future courses of policy. Such commitments allow the government to manipulate expectations, which are essential to determining incentives and consequently economic behavior.

However, there is a downside to expectations of higher future subsidies. These expectations may increase R&D, but they will also create an incentive for consumers to defer purchases into the future, whereas the government would prefer to shift consumption to today if possible. Moreover, there's a risk that, as the firms are realistically at least somewhat credit constrained, the loss of demand in the present may shrink the market and drive some firms out of business, reducing the supply of solar panels tomorrow.

Is it possible to subsidize purchases today more than purchases tomorrow, but to also subsidize R&D directly? Such a system could be set up by, for instance, pairing a higher subsidy of production today with a special investment tax credit. This would increase consumption of solar panels today while simultaneously stimulating research. However, this is a dangerous game, since both programs tend to counteract the other. The investment tax credit would lower consumers' expectations of future prices, which would cause them to shift consumption to the future. Likewise, increasing the current subsidy decreases the firm's expectations of future demand, in turn decreasing the expected return of R&D.

The above discussion demonstrates just how complicated a theoretically simple analysis of externalities may become, once you factor in the dynamic consumption and investment problem. I don't know whether the optimal subsidy is increasing or decreasing over time, and I don't know whether there should be separate subsidies for both research and production. To properly answer these questions, we would have to work out the problem I outline above in a little more detail, and then estimate the model parameters. I can only hope that California undertook such a dynamic analysis when it developed its subsidies.

How Governments make Markets

2009-08-09T22:37:00.028-04:00

The debate over economic policy is often framed as a question of governments vs. markets. Conservatives, the conventional wisdom goes, believe in less government and more markets at the margin, while liberals believe in more government and less markets.

While such discussion is informative, I think that it is in one sense flawed. Such a government-market distinction conceives of markets as natural features of human societies, at odds with government. However, my reading of human history is a bit different.

Markets come into existence when human beings interact by exchanging goods and services. Ideally, economic exchanges are mutually beneficial and wealth-creating. Yet this is not the natural state of interaction; rather, history shows that human interactions almost invariably devolve into zero-sum rent-seeking activities, and negative-sum destructive conflicts over fixed resources. Sadly enough, human history is not principally a story of steady material progress; it is a bloody tale of struggle over resources, with the warrior as hero, selling protection to peasants in return for their excess produce and making war on his neighbors to cement his power, augment his prestige, and justify his position.

Yet the potential for the betterment of all through exchange was too great to ignore. Wherever economic producers interacted, they spontaneously and sometimes intentionally developed social institutions to regulate their exchanges in order to eliminate negative-sum competitions and channel their energies into creative rather than destructive acts. As larger and more complex economies developed, these control systems became correspondingly intricate, until modern government in all its bureaucratic glory finally emerged.

Which brings us to the conflict between conservatives and liberals. Conservatives see a complex system of economic institutions that support our current "free" market economy. While almost all the separate pieces of this system were created by humans, there was no unifying plan behind the edifice. Each reform was undertaken piecemeal as regular people observed destructive rent-seeking behaviors, and pressured authorities to regulate the system in order to eliminate these. In the words of GMU's Bryan Caplan, conservatives "...believe that tradition incorporates the evolved use of information to solve problems..." which leads conservatives, "...to be very cautious about overthrowing existing institutional arrangements."

Liberals, by contrast, believe that the existing institutional arrangement is flawed, and trust experts to fix it. Again in the words of Caplan, they "...believe that expert technocrats should be in charge...[and] are comfortable with throwing out tradition and markets in order to cede power to experts."

Framing the debate in this manner is helpful because it abstracts from the question of "more" vs. less" government, and asks us whether we should undertake institutional reform in an incremental manner that respects the complex and diverse sources of the existing arrangement, or whether we trust our understanding of the system sufficiently to undertake radical change.

Needless to say, I have limited patience for those who appeal to romanticized views of an unregulated free market. Such may exist in isolated instances, but functioning markets are almost invariably regulated markets. The question is not whether markets should be free or regulated, or whether governments should be large or small. The key question is whether we sufficiently understand the current system, both its strengths and shortcomings, to undertake radical institutional reforms to correct its problems. I don't know the answer to this question, but it is definitely worthwhile to discuss.

Is Development Worth It?

2009-08-09T09:58:00.016-04:00

In a recent article in the New York Times, an Indian journalist reflects on the dramatic changes that his hometown has seen in the last 15 years. He strikes a balanced tone, discussing the benefits and costs of economic development one after another. For instance:

Development has for the most part been kind to this area. It’s created new wealth, new jobs, new opportunities. The men and women who have set up shops and restaurants and yoga centers on the road to the beach have done well for themselves.

But development has also disrupted existing ways of living. It has strained the social and cultural fabric of the villages. Kuilapalayam, a village at the head of the road leading to the beach, has had at least seven murders in recent years. Gangs of young men roam the village, extorting money, exacting revenge. Once, the panchayat, a traditional assembly made up of village elders, would have controlled the violence. But the new generation has modern ideas; they don’t heed their elders, and the panchayat membersare powerless, too scared to step in.

Development has led to new resentments and torn apart families. Farmers who used to toil over barren patches of land suddenly find that that land is worth a small fortune. They’ve built new houses, sent their kids to school, bought motorcycles and maybe even cars. A couple of universities up the road have widened people’s horizons.

But neighbors who didn’t own land, who watched their friends get rich, often don’t feel quite as sanguine about the changes. And long-forgotten relatives have appeared, perhaps returning from the cities to make a claim on the land. The papers are full of often violent stories about disputes over property.

Although I agree that economic development can often destroy things of value, I think we have to keep this in perspective. In 1911-1920, Indian life expectancy at birth was 20 years. By 1995, it had risen to 58.5 for men and 59.6 for women.[1] Today, it's up to 67.46 and 72.61 years respectively.[2]

Of course, a lot of these increases are due to declines in infant mortality. Between 1995 and today, infant mortality has fallen from over 76 deaths per 1,000 live births, to 30.15 per 1,000 live births.

While many poor remain, the World Bank estimates that the share of Indians living on less than the equivalent of $1 per day has fallen from 42% in 1981 to 24% in 2005.[3]

There are legitimate concerns about economic development. Growth often has negative externalities on others, and if those suffering are much poorer than those benefiting, the benefits would have to be correspondingly large to make it worthwhile. Yet India is so poor that economic growth that adds pennies to per capita income reaps benefits in infant mortality, life expectancy, and literacy, and it's unclear who else is suffering from this growth beyond a few manufacturing workers in the rich nations.

What about the destruction of traditional modes of life in India itself? It's difficult to measure such destruction, and more difficult still to assign a value to it. Yet when we're discussing 10 extra years of human life per person, across 1.15 billion people, there would have to be an awful lot of destruction of traditional values to even out the balances.

Unfortunately, there is a market in the US for the perspective that the New York Times is selling. It's "cool" to be at least a little skeptical of economic growth, to shed a tear for the romantic past swept away in the rush of crass consumerism. This stems in part from legitimate critiques of the material excesses of American culture, but it is woefully misplaced when applied to developing countries like India. There is no moral equivalence between 11.5 billions years of human life, and disrespect for traditional authorities.

Bad News on Business Conditions

2009-07-02T22:20:00.004-04:00

As a follow up to my last post....

The new employment report that came out today was very bad news:

Before this last report, you could be forgiven for thinking that the labor market was rapidly bottoming out. Note that the chart above is the change in employment, so it's already a first derivative. Anything below 0 means that things are getting worse. Or more accurately, anything below the number needed to absorb new workers entering the workforce, so no change would need to be around 200,000 new jobs per month. However, the graph above is a major bump in the road back to 200,000.

Inflation expectations look like pretty much zilch right now, which isn't good either. Despite a ton of quantitative easing, the Fed hasn't managed to gain any traction, and there's talk of keeping interest rates at zero for a long time to come. Not good.

Good News on Business Conditions

2009-06-27T00:41:00.005-04:00

There's been a lot of discussion recently about whether things are looking up. Here's a piece of evidence in the positive direction. This is the ADS index:

The ADS index is based on a paper by Francis Diebold (Penn), Aruoba (Maryland), and Scotti (Fed). It purports to track business conditions in real time. The shaded area corresponds to the officially designated recession. Note that the beginning of the recession was evident in real time in the index, although it was officially declared retrospectively. However, the index has been steadily climbing since January, and has almost reached the level it held in September 2008, when the situation began to deteriorate rapidly.

For historical perspective, here's a longer version of the same chart, from ADS's original paper:

This Doesn't Make Sense to Me...

2009-06-24T13:27:00.011-04:00

Paul Krugman, in a post entitled The President is Making Sense, quotes Obama:

QUESTION: Wouldn’t that [a public health insurance option] drive private insurance out of business?

OBAMA: Why would it drive private insurance out of business? If private insurers say that the marketplace provides the best quality health care; if they tell us that they’re offering a good deal, then why is it that the government, which they say can’t run anything, suddenly is going to drive them out of business? That’s not logical.

Nobody is claiming that markets offer the best quality; rather, markets offer the most efficiency. Companies don't compete on quality alone; rather, they compete on both quality and price. The theory is that private companies will offer better quality than a public company if they are subject to the same constraints. If the government operates a public company without making a profit, or even at a loss, then it will have a price-advantage over private companies that will drive them out of business, even if the public company is less efficient than the private ones.

This isn't to say that the health care system doesn't need to be reformed, but that creating a public option is certainly not the right way to go about it.

The Six Questions of Economic Growth

2009-06-23T14:35:00.009-04:00

As I consider The Most Important Question in Economics, I thought it would be useful to list the basic facts of economic growth, and the corresponding questions.

First, the basic facts:

(1) Before around 1800, the average welfare of human beings in all human societies was approximately equal, and tended towards subsistence level. Any growth in the productive capacity of society was eventually absorbed by population growth, leaving human beings in England in 1800 no better off than their ancestors 5000 years before.

(2) Starting around 1800, there was rapid and accelerating growth in living standards beginning in England and spreading quickly to most of Western Europe and North America.

(3) Throughout the 20th Century, a number of additional countries have undergone phases of rapid economic growth, which has set them on a course from subsistence level to convergence with the West. These countries include parts of Southern and Eastern Europe, Japan, Korea, and a number of other East Asian countries, and in recent years has include countries such as Brazil, India, China, and Bangladesh.

(4) Despite these "growth miracles", many countries remain mired in extreme poverty, and have even seen their standards of living fall in real terms throughout the past century.

(5) Deterioration of certain measures of the environment, both on the global and local levels, have demonstrated the public costs of economic wealth, and it is unclear whether the currently wealthy countries will be able to sustain their economic growth, or even their current living standards, particularly if the developing world catches up to them in living standards.

(6) Moreover, many wealthy countries have seen large income gaps open up between certain sub-populations, whether defined by ethnicity, class, or geography.

This leads to six questions:

(1) Why was their no economic growth before about 1800?

(2) Why did economic growth begin where it did and spread rapidly to certain countries, but not to others?

(3) Why have certain countries begin been able to undergo periods of rapid growth that have allowed them to converge with the West, while others remain mired in poverty?

(4) How can those countries that remain mired in extreme poverty enter into sustained economic growth.

(5) How do we address environmental degradation, and can we sustain growth in the future, particularly as developing countries catch up with the West?

(6) Why hasn't economic growth lifted all boats in the developing world, and how can we stimulate growth among sub-populations that remain impoverished in the midst of plenty?

As a Christian, I must also struggle with how to explain these facts from a Gospel perspective. It seems clear that the last 200 years are materially unprecedented in human history, and I am uncertain as to how to synthesize them with a biblical worldview. But more on this seventh question later.

The MPG Illusion

2009-05-22T10:57:00.013-04:00

As long as I'm beating up on CAFE standards, I might as well point out the somewhat deceptive nature of MPG as an efficiency standard. MPG gives the number of miles that can be driven using 1 gallon of gasoline. However, a more intuitive metric would be gallons/mile, or perhaps gallons per 1000 miles (just so we don't have to use small decimal numbers).

Suppose that the average mileage was 10 MPG. This corresponds to 100 g/kmi. Conversely, if the average mileage was 100 MPG, we would have 10 g/kmi. This has a certain symmetry, so you might expect linear scaling between the two. So does 55 mpg correspond to 55 g/kmi?

Nope. If you get 55 miles per gallon, it only takes you 18.2 gallons to get to 1000 miles. This is because mpg is a geometric measure, while g/kmi is linear. Going from 10 mpg to 55 mpg means that I have made my car 5.5x as efficient, and to get that same boost in efficiency again, I would have to go up to 303 mpg.

Of course, human beings tend to think in linear terms. So when we think about three cars with 20, 30, and 40 mpg respectively, we think that the third car is as much better than the second as the second is better than the first. This is not true. The first car takes 50 gallons to go 1000 miles, while the second takes 33 gallons, and the third takes 25. In other words, we would save about 17 gallons per 1000 miles by using the second car over the first, while we would save only an extra 8 gallons by using the third over the second. The second 10 mpg are worth a lot less than the first 10 mpg.

Which brings me to the new CAFE standards proposed by the administration. Consider four numbers. If mpg remains constant, we would have 26 mpg in 2016. Under the rules put in place last year by the Bush administration, we would have about 32 mpg. Under the new standards proposed by the Obama administration, we would have about 35.5 mpg, and if technology we pursued to its fullest, we might be able to get up to 42 mpg, so <= 42 mpg is the feasible set.

Let's convert these numbers into gallons per 1000 miles.

Baseline: 38 g/kpmi
Bush: 31 g/kpmi
Obama: 28 g/kpmi
Feasible: 24 g/kpmi

In other words, the best feasible mileage under current technology would save us about 14 gallons per 1,000 miles driven, which would reduce gasoline consumption by about 37% per mile driven. The Bush standard realizes about half of these feasible gains, saving about 7 gallons of gas per 1000 miles, or 18%. The Obama standard saves an extra 3 gallons for total savings of 10 gallons per 1000 miles, or 26%.

So there you have it. Obama's new standards will save us about 3 gallons of gasoline per 1,000 miles over the current schedule, or 10 gallons relative to current standards. In percentage terms, the Obama plan would save us about 8% extra for each mile driven more than the Bush plan.

CAFE Standards are Stupid

2009-05-19T13:37:00.002-04:00

As economic policy, setting fuel efficiency standards is stupid. It's not stupid because it will have no effect, but because it is a strictly inferior policy to a very simple alternative -- increasing the gasoline tax.

What happens when the government mandates higher fuel efficiency standards? Applying these standards is costly, so this amounts to a tax on the production of cars. Since tax incidence is independent of where the tax is applied, this is equivalent to a tax on the purchase of cars. Therefore, cars will be more expensive, and fewer will be bought and sold. Due to the higher expense of purchasing a car, people will retain old cars longer, incurring additional inconvenience and repair fees rather than pay a higher price for a new car.

Of course, those new cars that are bought will be more fuel efficient on average. Because the average cost of driving to car-owners has decreased, people will drive more. Since both miles driven and miles per gallon will increase, it is unclear whether the total consumption (and hence price) of gasoline will increase or decrease, but accidents and congestion will definitely increase, and the use of public transit and other alternatives to driving will decrease.

This is because mandating higher fuel efficiency standards is the equivalent of taxing the purchase and manufacture of cars while subsidizing their use. What we want is to tax the use of cars, not their manufacture. This can be done through a gasoline tax increase.

If we raise the tax on gasoline, the price of gasoline will increase. In response, people will economize on their use of gasoline by purchasing fuel-efficient vehicles, reducing trips that require driving, and substituting other means of transportation. New car purchases may increase as people swap older inefficient cars for newer efficient cars. This policy will have the additional benefit of decreasing other negative externalities of driving, such as congestion and accidents.

So why is the Obama administration proposing stricter CAFE standards? Politics. "Gasoline tax" sounds scary. "Better standards" sounds nice and friendly -- who doesn't want manufacturers to abide by good standards? Yet it's nothing more than a hidden tax, and in their haste to hide the tax, the administration has chosen to apply a very inefficient one.

Could The Government Benefit from some Competition?

2009-04-10T09:33:00.002-04:00

Arnold Kling writes:

Think of government as a charity. From a libertarian perspective, it is a charity run by the Mafia, which will break your knuckles if you don't make your donations. It is also a badly mismanaged charity. It funnels lots of money into questionable causes, and even when the causes are good the programs that it funds tend to be very wasteful.

I would like to see government have to compete with other charities on a level playing field. I often say that government should have had to fund the financial bailouts by holding a bake sale.

How might we set up such competition? One way would be to let donations to a certain class of charities be 100% tax deductible, meaning that you can decrease your tax liability by the full amount of your donation, up to a certain portion of your total liability.

Thus if people think that a charity will do more good at the margin than government, they can pay some of their taxes to that charity instead of to the government. If you like the idea of universal health care, you could contribute 50% of your taxes to a chain of free clinics.

The Great Ricardian Equivalence Misunderstanding

2009-04-08T12:26:00.004-04:00

Ricardian equivalence is a concept in economics. It shows that, under certain assumptions, funding a given level of government expenditures via borrowing is equivalent to funding them via taxation.

Paul Krugman argues that, even in a world with perfect Ricardian equivalence, a temporary government stimulus can boost demand, although a permanent increase cannot.

Paul Krugman:

Here’s what we agree on: if consumers have perfect foresight, live forever, have perfect access to capital markets, etc., then they will take into account the expected future burden of taxes to pay for government spending. If the government introduces a new program that will spend $100 billion a year forever, then taxes must ultimately go up by the present-value equivalent of $100 billion forever. Assume that consumers want to reduce consumption by the same amount every year to offset this tax burden; then consumer spending will fall by $100 billion per year to compensate, wiping out any expansionary effect of the government spending.

But suppose that the increase in government spending is temporary, not permanent — that it will increase spending by $100 billion per year for only 1 or 2 years, not forever. This clearly implies a lower future tax burden than $100 billion a year forever, and therefore implies a fall in consumer spending of less than $100 billion per year. So the spending program IS expansionary in this case, EVEN IF you have full Ricardian equivalence.

Is that explanation clear enough to get through? Is there anybody out there?

I disagree. Krugman is forgetting that in a Ricardian equivalence world those receiving government funds will save in order to smooth their consumption as well. I'm going to use a very simplistic model to demonstrate this.

Consider a two period model. In case 1, the government increases spending by g today and g tomorrow. In present value, the government increases spending by g + g/(1+r) = g*(2+r)/(1+r). The consumers realize that government will have to raise taxes in the following period to pay for these expenditures. In particular, government will have to pay down this period's expenditure of g, plus interest on that g, plus g tomorrow. So consumers anticipate a tax next period equal to g*(2+r). Because they want to smooth consumption, they will save g dollars, so that next period they will have g*(1+r) in the bank to pay down the tax, and their consumption will fall by g in each period. Thus there is no increase in expenditures in case 1.

Now let's consider case 2. Government increases expenditures by g in this period ONLY. Thus, the net present value of today's expenditures are g, and taxpayers anticipate a tax increase of g*(1+r) next period. If they want to smooth consumption, how much will they save? Suppose they save s. Next period, they will have (1+r)*s to pay the tax of (1+r)*g, plus their contribution in period 2, which we shall call d. Since they want to smooth consumption, d = s. Therefore, (1+r)*g = (1+r)*s + s, and so s = g*(1+r)/(2+r). Since g*(1+r)/(2+r) < g, this might appear to be expansionary.

However, let's not forget about the people receiving the government money! In case 1 we can be forgiven for forgetting about them: they receive g in each period, and so they spend g in each period. However, in case 2 they receive g in period 1 only. They want to smooth their consumption, so they save an amount s. Next period, they enjoy (1+r)*s additional consumption, which must be equal to whatever they kept last period. Thus g - s = (1+r)*s, so s = g/(2+r). In other words, they increase their consumption by g - s = g - g/(2+r) = g*(1+r)/(2+r). Which is exactly equal to the amount by which the tax-payers decreased their consumption.

Therefore government spending is never expansionary in a Ricardian equivalence world, whether it's temporary or permanent, debt-financed or tax-financed.

Now you might wonder whether this applies to the case in which government spending takes the form of building a bridge, rather than just giving the money to people. You also might wonder to what degree Ricardian equivalence holds in our world. I'll tackle those questions in a future post.

A Commentary on the Geithner Plan

2009-04-02T16:10:00.002-04:00

The Geithner bank recapitalization plan has been excoriated by economic commentators for its implicit subsidy of the purchase of legacy assets. The plan partners with private investors, providing them with a large non-recourse loan to cover 85% of the purchase of legacy assets. Because the loan is non-recourse, if the asset ends up being worthless, the private investor is on the line for only his contribution. Paul Krugman:

Leave on one side the question of whether the Geither plan is a good idea or not. One thing is clearly false in the way it’s being presented: administration officials keep saying that there’s no subsidy involved, that investors would share in the downside. That’s just wrong. Why? Because of the non-recourse loans, which reportedly will finance 85 percent of the asset purchases.

Let me offer a numerical example. Suppose that there’s an asset with an uncertain value: there’s an equal chance that it will be worth either 150 or 50. So the expected value is 100.

But suppose that I can buy this asset with a nonrecourse loan equal to 85 percent of the purchase price. How much would I be willing to pay for the asset?

The answer is, slightly over 130. Why? All I have to put up is 15 percent of the price — 19.5, if the asset costs 130. That’s the most I can lose. On the other hand, if the asset turns out to be worth 150, I gain 20. So it’s a good deal for me.

This plan therefore constitutes an implicit subsidy. Detractors have argued that if the problem was simply liquidity (i.e. that private actors couldn't get the financing to purchase these assets in bulk), the government could have provided financing to private actors without a subsidy. The fact that such a subsidy is necessary implies that the assets are worth less than what will ultimately be paid for them, and the taxpayers will be on the hook for the difference. Since this is the equivalent of just giving money to banks (direct recapitalization), it would be much better if the government simply took over and recapitalized the bad banks.

While this analysis may well be correct, there is another way of looking at this problem. Suppose that purchasing questionable assets has a positive externality for the whole economic system. Standard economic theory suggests that activities with positive externalities should be subsidized.

Consider the following modified version of Krugman's exmaple. Again suppose that the asset is worth either 50 or 150. If the economy is in recession, the odds are 50/50 either way. But suppose that if the economy is in recovery, the odds are 90/10 that the asset will be worth 150. Moreover, the probability of recover is a function of whether these assets are taken off the bank's balance sheet.

Suppose that an investor thinks that the probability of recovery in the next year is 10%. Then he will value the asset at (0.10)(0.10*50 + 0.90*150) + (0.90)(0.5*50 + 0.5*150) = 104. If everyone buys up the assets, the probability of recovery becomes 90%, so the value of the asset in this case is (0.90)(0.10*50 + 0.90*150) + (0.1)(0.5*50 + 0.5*150) = 136. Suppose the bank will not sell the asset for less than 120, and so the market is frozen.

Suppose our investor participates in the Geithner plan. If he pays P for the asset, he will make 150 - P if he wins, and -0.15P if he loses. If he pays 132, he makes 18 if he wins and loses about 20 if he loses. Therefore his expected profits are:
(0.10)(0.9*(18) - 0.1*(20)) + (0.90)(0.5*(18) - 0.5*(20)) =~ 0.5

Thus he will pay any price less than 132. Since this is higher than the bank's reservation price of 120, the markets will become unfrozen and the banks will sell to the investors at a price of around 128. Because the money is flowing, we will hopefully move to the p(recovery)=0.9 world, in which the assets are worth 136. Thus the government will on average make money on these assets.

Of course, I chose these numbers in order to make my example work out nicely. The Geithner subsidy might be too large, and the government might lose money on it. The larger point is that subsidizing purchases of these assets could be a perfectly reasonable policy if we believe that there is coordination failure in the market for these assets.

Incidentally, the debate over the Geithner plan is a case of a fairly fundamental philosophical question that I have discussed preciously on this blog: if there is a market failure in the provision of a service, should the government try to fix the market for the service, or should the government provide the service itself?

I hope you all know where I come down on this question.

Telling Stories

2009-03-27T16:53:00.013-04:00

We economists are clever creatures. If you give us an empirical result, we can almost always spin a tale to explain it. Even if the result is unintuitive and contradicts our initial predictions, we can almost always come up with a story to explain it ex post facto.

This seems to suggest a weakness in economic thinking. After all, if you're able to explain all outcomes equally well, then you aren't really explaining anything at all. A mental model that accurately reflects reality should be baffled by non-truth. Having a clever answer ready at a moment's notice suggests weakness, not strength.

Yet in many ways this is unavoidable. Economists are trying to model phenomena that are the product of largely unobservable factors interacting in nonlinear ways. Economic models don't try to be simplifications of reality, rather than exact replications as in the hard sciences. Thus economics is a process of choosing among various simplifications of reality designed to capture the sorts of interactions relevant in the particular class of cases under consideration. In this context, a good economic intuition for story-telling is a valuable asset.

However, it's very important that economists don't just tell stories and leave it at that. The purpose of economic story-telling is not to explain away reality, but to propose testable explanations, which can lead to the formulation of a model more appropriate to the particular problem.

One big risk here is that economists may fall in love with their own cleverness, preferring their own original counter-intuitive explanations when a more commonplace model would work just fine. This is why I think that it is helpful for economists to assign explicit probability estimates to various stories, and make sure that they have considered the full universe of possibilities. Most papers include a literature review describing the state of knowledge on the subject, but few researchers present priors based on this information, nor explicitly describe how their results should shift these distributions.

Rosy CEA Projections and the Winner's Curse

2009-03-23T10:42:00.007-04:00

The Winner's Curse is a concept in game theory (auction theory, really). It states that, when multiple parties bid on an asset of uncertain objective value, the winner will generally bid more than the asset is worth.

To illustrate this, imagine 10 oil companies bidding for an oil field. Each is allowed to assess the oil field's value by drilling in one location. If oil is distributed unevenly throughout the field, we might model the amount of oil extracted from any given well as the mean quantity of oil plus an error term.

Some of the companies will be "lucky" and hit more oil. These companies will assess the field's worth more highly, and thus bid more for the field. The company that enters the winning bid is probably one of the "luckiest" companies, and has most likely overestimated the true value of the field.

I propose that a similar problem exists for policymakers, a "policymakers' curse" if you will. Suppose that there are ten possible policies under consideration. The CEA forecasts the economy under each of the possible policies. Suppose that the CEA is an unbiased forecaster. Then we can model each forecast (of GDP growth, for instance) as a random variable with mean equal to the true GDP growth of pursuing that policy, plus an error term. If policy-makers decide to pursue the course of action with the best assessment, it's very likely that they have chosen a course of action with a highly positive error term.

In other words, if some of the CEA's forecasts are more optimistic, and some are less optimistic, the forecasts with the "best" result is likely to be an optimistic one. We don't need to assume a biased forecasting process to conclude that official projections are likely to be biased upwards. We only need to assume that policymakers take economic projections into account when choosing economic policy.

This prediction is borne out in the recent mini-controversy surrounding the "rosy" CEA projections for the economy's future. Since these projections were not independent of the policy-making process, they likely overestimated future economic growth under their own recommended policies. As one might expect, the CEA's projections are more optimistic than most independent forecasters and the CBO's estimates.

What can be done to offset this effect? We can draw a lesson from the practices of company's in auctions. Since participants in auctions have realized that there is a correlation between the event of their winning and the event of their overestimating the true value of the asset, they have began assuming that their estimate of the value of the asset is biased upward. Thus they have started marking down their bids. I suggest that we do the same when considering the CEA projections.

A Simple Explanation of the Great Recession

2009-03-12T21:14:00.005-04:00

Much ink has been spilled trying to boil down the diverse causes of the Great Recession to the original proximal cause. This is my attempt:

A credit default swap (CDS) is a kind of insurance against credit risk. It is a privately negotiated bilateral contract. The buyer of protection pays a fixed fee or premium to the seller of protection for a period of time and if a certain pre-specified “credit event” occurs, the protection seller pays compensation to the protection buyer....

CDSs can...be used as a way to gain exposure to credit risk. While the risk profile of a CDS is similar to a corporate bond of the reference entity, there are some important differences. A CDS does not require an initial funding, which allows leveraged positions. Moreover, a CDS transaction can be entered where a cash bond of the reference entity of a particular maturity is not available. Further, by entering a CDS as protection buyer, one can easily create a ‘short’ position in the reference credit. With all these attributes, CDSs can be a great tool for diversifying or hedging one’s portfolio.

(Nomura Fixed Income Research, 2004)

Imagine an insurance company that insures a lot of homes in Florida. Suppose the company sets premiums such that it makes a profit on average, based on the individual probabilities of any given home being damaged or destroyed.

However, the probabilities of Florida insurance policies being called in are not independent. Suppose there is a massive hurricane that destroys millions of homes in Florida. Our unfortunate insurance company will face billions of dollars in losses. If the company had not anticipated such an event, it very likely will go bankrupt.

In this case, the damage from the hurricane to the financial system will be equal to the physical damage -- say 1 billion dollars. However, suppose that many other companies had placed bets based on these contracts being called in. Since there is no limit on such bets, there could be hundreds of billions of dollars worth of payments that would suddenly come due in the event of a hurricane. This would represent a massive shock to the economic system, likely causing widespread bankruptcies. Such a catastrophe would be very likely to kick off a spiral of Fisherian asset deflation, that could cripple the economy.

This is just what happened in the US. Credit default swaps allowed financial institutions to place bets that would come due in the event of a default on an underlying asset. The hurricane was a general decline in housing prices. When this took place, it became clear that the probabilities of default on a wide range of mortgage-backed assets were correlated.

Although the total decline in "value" of the housing stock was fairly small compared to the US economy, the excessive leveraging allowed by CDSs left the financial system strangely vulnerable to just such a decline. The results have not been pretty.

There have been many other factors contributing to the present crisis. However, if I were asked to give a simple explanation based on the single greatest proximal cause, CDSs would be it.

Response to Brad DeLong

2009-02-24T18:14:00.031-05:00

Brad DeLong:

[T]here is a surprisingly large current of thought stating that stimulus packages simply do not work, ever....I do not understand their argument.

Back at the start of 1996, the U.S. unemployment rate was 5.6%. Then the assembled investors and businesses of the United States discovered the internet. Over the next four years spending on information technology equipment and software in the United States roared upwards from a pace of $281 to a pace of $446 billion a year—and the U.S. unemployment rate dropped from 5.6% to 4.0% and the economy grew at a 4.3% real annual rate as the spending of the high-tech boom pulled extra workers out of
unemployment and into jobs.

Back at the start of 2004 America’s banks discovered that they could borrow money cheaply from Asia and lend it out in higher-yielding domestic mortgages while using sophisticated financial engineering to...wall off and strictly control their risks while making lots of money. Over the next two years spending [on] building residential structures in the American economy roared upward from a pace of $624 to a pace of $798 billion a year—and the U.S. unemployment rate dropped from 5.7% to 4.6% as the economy grew at a 3.1% annual rate.

In both these cases large groups of people in America decided to increase
their spending... [and] their increased spending...pull[ed] large numbers of Americans...into productive and valued employment.

The government’s money is as good as anybody else’s. If businesses’ enthusiasm for and spending on high-tech and if new homeowners’ enthusiasm for and spending on 3 BR, 2BA can boost employment and production...then what is the argument that...government spending cannot boost employment and production? I just do not see it.

Bryan Caplan responds by pointing out that the money for spending must come from somewhere, and if it hadn't been spent on that it would have been spent on something else. However, I don't like this thinking because it focuses on the wrong part of the equation. Money is a medium of exchange, but it doesn't have real effects unless there's either too little or too much relative to expectations. Talking about "where do the funds come from?" is secondary, because the key isn't the funds. The key is the resources.

The economy at any given instant has only so many productive resources. Some of these resources are idle, while some are employed in productive enterprise. To the extent that any action affects the "economy", it does so by moving these resources from one use to another, or else bringing idle resources into employment.

So what stimulated the internet boom? The internet boom took place when a lot of actors in the economy anticipated higher productivity in the tech sector. This lead to a lot of spending in that sector, which had the effect of shifting resources from other uses into the tech sector. This lead to economic growth.

We can break down this economic growth into three categories: productivity gains, employment gains, and coordinative effects. To the extent that people were right that resources were more productive in the tech sector than wherever they had been before, there were real gains in income due to moving those resources into a more productive sector. Furthermore, since some formerly unemployed or idle resources may have been brought into play by the expectation of higher returns, some of this growth was the result of net employment growth. Finally, expectations of economic growth lead to coordinative investment, the so-called "virtuous cycle" of economic growth.

Now let's consider an increase in government spending in the form of a stimulus bill. I think that we can all agree that government spending is unlikely to move resources into high productivity sectors better than the market. Even most stimulus advocates agree that it's reasonable to mark down the value of the actual spending to something less than the ticket price, although some will argue that widespread under-investment in needed public goods means that government spending may be worth its price, or even exceed it.

However, most stimulus advocates are banking on the employment and coordinative effects to really make the difference. The employment effect is likely to be large in the short-run, since there is a lot of cyclic unemployment in the economy at the moment. However, in the long-run the stimulus plan will likely cause large-scale crowding out in the labor market, which will drive up wages -- just the opposite of what we want.

The more important avenue is the coordinative effect. The idea is that the short-run jolt to employment will convince private players that things are turning around, and will lead them to start expanding as well. Ideally, this will reverse the current vicious cycle of economic decline and begin the virtuous cycle of economic growth.

DeLong's use of the '90s tech boom as an argument in favor of government spending suggests that he views the economic growth of the '90s as largely the consequence of coordinative effects. If this is true, than it didn't really matter that it was tech. If the government had randomly picked a sector of the economy off the wall in the mid '90s and started shoveling money at it, we would have seen the tech boom all over again.

I disagree with this narrative. I believe that most of the economic growth in the late '90s can be attributed to productivity gains in the tech sector. It's true that investment overshot a bit due to over-employment and coordinative effects, but the correction was relatively mild. The most lasting damage from that recession may have been that we confused the sectoral shift from increasing to decreasing labor market participation for a weak labor market recovery, which lead to excessively loose monetary policy in the subsequent years. What role that played in our current crisis I cannot say for sure, but it certainly did not help.

Although I support the stimulus bill, I do not find Prof. DeLong's argument persuasive. I view economic growth as determined largely by short-run expectations and long-run productivity and demographics, not by coordinative effects. The case for reversing coordinative effects via stimulus rests on the idle resources in the economy, which is purely a short-run phenomenon.

My view is that the current recession is the consequence of an expectation shock to the economy, which has caused a large amount of cyclic unemployment and has launched a vicious cycle of economic decline. The tech and housing booms were entirely different, and reference to them is not useful.

My Contribution to Macroeconomics

2009-02-05T10:57:00.003-05:00

In classical pre-Keynes economics, the economic adjustment process was modeled as follows:
1. Economy is in equilibrium
2. Exogenous shock hits economy
3. Prices adjust in response to this shock
4. Economic decisions adjust in response to price changes

If you assume that this all happens instantaneously, then you get real business cycle theory -- that is, the ups and downs of the economy are the result of optimal adjustments in response to real economic conditions. The economy never runs into a shock that it can't handle optimally. Sometimes a shock will make us all much worse off, but that's because the shock was bad, not because it knocked the economy off its moorings.

The big contribution of Keynes was his insight that the adjustment process in the economy is not necessarily smooth. If prices do not adjust quickly, or if people do not rationally change their behavior in response to prices, then the market economy can fail to properly adjust to a shock.

One way people model this is by assuming that prices are fixed in the short run, but perfectly flexible in the long-run. These are the neoclassical synthesis models with "price stickiness".

I don't like these models. I'm working with an alternative model derived from proper micro foundations.

The key insight of my model is that price flexibility is costly. Economic agents choose how much flexibility they want in setting prices, and they will make that decision based on their expectations of price variations.

If price flexibility has positive externalities, then it follows that markets are generally less flexible than is optimal. A further implication is that markets are better at absorbing anticipated shocks than unanticipated shocks.

Is Obama's Honeymoon Over?

2009-02-04T17:12:00.003-05:00

Tyler Cowen thinks so:

Did the stimulus bill just fail?

Sort of, or so says The Washington Post. Apparently some of the moderate Democrats are not on board and there is talk of cutting it by as much as $200 billion.... Overall I am surprised that the bill didn't sail through. I expected one round of opposition and then a collapse as extra goodies were put on the Xmas tree for recalcitrant legislators. The political equilibrium now seems trickier than that and I believe that, for better or worse, Obama's "honeymoon" period is now over.

Paul Krugman lends credence to the notion, as he minces no words:

You know, it was widely expected that Obama would have a stimulus plan ready to pass Congress even before his inauguration. That didn’t happen. We were told that this was because the economic team was working flat out on the financial rescue.

In fact, when it comes to bank rescue it’s hard to see much evidence that anything was accomplished during all that time; the team is still — still! — running ideas up the flagpole to see if anyone salutes. And the ideas look remarkably bad....

Meanwhile, when it came to stimulus legislation, when Obama finally introduced his economic plan he immediately began negotiating with himself, preemptively offering concessions to the GOP, which voted against the plan anyway....

As a wise man recently said, failure to act effectively risks turning this slump into a catastrophe. Yet there’s a sense, watching the process so far, of low energy. What’s going on?

EDIT:
My personal view is that Obama is a politician and knows what he's doing. The reason he started compromising from day 1 is that he understands that swift fiscal action is difficult in the American political system. Obama simply cannot get what he wants and get it quickly. In fact, very often presidents fail to get either. I think that Obama recognizes that getting as much as possible as quickly as possible requires lots of compromise, and that it only can happen so fast.

Why do the rich spend less?

2009-02-03T12:52:00.003-05:00

In my previous post, I extensively quoted Robert Reich's analysis of the American economy of the past several years. I highlighted one section that I thought summed up our fundamental differences:

But the rich don't spend as much of their income as the middle class and the poor do -- after all, being rich means that you already have most of what you need. That's why the concentration of income at the top can lead to a big shortfall in overall demand and send the economy into a tailspin.

I've already expressed my disagreement with the aggregate demand argument, but I also want to take issue with Dr. Reich's explanation for why the rich save more. Simply put, his explanation does not accord with standard consumer theory.

Robert Reich seems to be arguing that the rich value marginal consumption less than the poor, which leads them to save more of their marginal income. This is incorrect. Standard theory implies that, if one values marginal consumption less (relative to leisure), than one will choose to work less. The saving-spending decision is based on how much one values present consumption versus future consumption.[1][2]

Given this framework, there are three reasons that I would expect rich people to save a higher share of their income than poor people:

1. Consumption smoothing: According to standard theory, there are declining returns to consumption (i.e. the first $1000 you consume is more enjoyable than the second $1000). Thus people with variable incomes throughout their lifetime will save some money when they're making more, and save less (or even borrow -- i.e. save a negative amount) when they're making less. Thus people in their 20s and 30s may save relatively little, since they expect to make more money later. When they're in their 40s and 50s, they will be making more money (and therefore richer), and they will be saving for retirement, and paying down debt/loans from their younger years. This is probably the most commonly given explanation for higher savings rates among the wealthy.

2. Time preferences: People generally value future consumption less than present consumption. Some people value the future at a relatively small discount relative to the present, while others discount the future by more. One might expect people who discount the future less to generally be more successful in life, because they will procrastinate less, work harder, and save more and therefore have more money when they're older. In other words, it's not that rich people tend to become thrifty, it's that thrifty people tend to become rich. Alternatively, it may be that wealthier people expect to live longer (Due to better health), and so save more for that future.

3. Higher expected returns: Wealthier people may be able to take advantage of economies of scales in investments, and they may also be better able to evaluate investment opportunities on average. In other words, wealthier people are on average likely to be better at managing money (that's how they became rich!), and so will expect higher average returns for their savings, leading them to save more.

This is highly significant for policy, because any program that transfers wealth from the rich to the poor (or vice versa) will affect the economy's savings rate. If you're a neoclassical economist like me (and like many on Obama's staff), then you think that a higher savings rate makes a nation wealthier. If you're a Keynesian, then you think that a higher savings rate makes a nation poorer. Here is an empirical look at the question.

Notes:
1. One might argue that present consumption of luxury goods may be constrained relative to future consumption. However, one can model this as deflationary expectations, which increases real returns of savings, and so is covered by point 3.
2. Savings can also represent present consumption if they are intended to be passed on to one's heirs. For instance, a rich person might save money planning to pass it on to his or her loved ones, and imagining their future happiness would be pleasant for him/her in the present.

Cyclists vs. Structuralists

2009-02-01T14:07:00.003-05:00

I support the stimulus bill. Personally, I would prefer a bill that was bit more focused, and possibly a bit smaller, but I think that something very much like it should be passed.

It looks like the final stimulus bill, hopefully a somewhat improved version, will be passed into law sometime in February. Most economic forecasters predict that, with the help of the bill, the economy should begin to improve sometime late this year, and should be substantially recovered by 2012 at the latest.

The bulk of the spending in the stimulus plan will end around that time. What happens next will be very interesting. In the words of Robert Reich:

...the moment the economy appears to be on the mend, conservatives such as Feldstein will want the government to cut spending. In their view, this is the only way to get the economy fully back on track. But others believe that it is precisely the track we were on that got us into this mess in the first place.

Those who support the stimulus as a desperate measure to arrest the downward plunge in the business cycle might be called cyclists. Others, including me, see the stimulus as the first step toward addressing deep structural flaws in the economy. We are the structuralists. These two camps are united behind the current stimulus, but may not be for long. Cyclists blame the current crisis on a speculative bubble that threw the economy's self-regulating mechanisms out of whack. They say that we can avoid future downturns if the Fed pops bubbles earlier by raising interest rates when speculation heats up.

But structuralists see it very differently. The bursting of the housing bubble caused the current crisis, but the underlying problem began much earlier -- in the late 1970s, when median U.S. incomes began to stall. Because wages got hit then by the double-whammy of global competition and new technologies<, the typical American family was able to maintain its living standard only if women went into the workforce in larger numbers, and later, only if everyone worked longer hours.

When even these coping mechanisms were exhausted, families went into debt -- a strategy that was viable as long as home values continued to rise. But when the housing bubble burst, families were no longer able to easily refinance and take out home-equity loans. The result: Americans no longer have the money to keep consuming. When you consider that consumers make up 70 percent of the economy, the magnitude of the problem becomes apparent.

What happened to the money? According to researchers Thomas Piketty and Emmanuel Saez, since the late 1970s, a greater and greater share of national income has gone to people at the top of the earnings ladder. As late as 1976, the richest 1 percent of the country took home about 9 percent of the total national income. By 2006, they were pocketing more than 20 percent. But the rich don't spend as much of their income as the middle class and the poor do -- after all, being rich means that you already have most of what you need. That's why the concentration of income at the top can lead to a big shortfall in overall demand and send the economy into a tailspin. (It's not coincidental that 1928 was the last time that the top 1 percent took home more than 20 percent of the nation's income.)

Other structural problems are growing as well. One is climate change and our dependence on oil. Another is the United States' growing reliance on foreign capital, mostly from China, Japan and the Middle East. Neither is sustainable.

Meanwhile, our broken health-care system drains more of our dollars yet delivers less care. When President Clinton tried to tackle health care in 1994, it represented 14 percent of our GDP, and 38 million Americans were uninsured. Now, the nation spends 16 percent of its GDP on health, and about 44 million of us are uninsured. Most cyclists acknowledge these problems, but they tend to think of them as separate from the current crisis -- issues to be tackled after the economy has recovered, and then only to the extent that we can afford to do so.

But structuralists like myself don't believe that the economy can fully recover unless these underlying problems are addressed. Without policies that put the nation on the path to higher median incomes, higher productivity, renewable energy and a more accessible and efficient health-care system, we'll face deeper and more prolonged recessions, followed by ever more anemic upturns. Bill Clinton's inability to do enough about these problems in the 1990s, followed by George W. Bush's negligent disregard of them, allowed them to grow to the point where any major triggering event can cause a vicious downward spiral.

As early as next year, the business cycle may hit bottom and begin climbing. At that point, cyclists and structuralists will want two different things -- and which side the president chooses will be, as Will Marshall of the Progressive Policy Institute puts it, the "central drama" of the Obama administration. The president recently sought to placate the cyclists by promising to focus on controlling the future costs of Social Security and Medicare. Perhaps Obama has in mind a "grand bargain" that would allow him to continue his structural agenda after the business cycle turns upward in return for limiting spending for those two giant entitlement programs.

Let's hope he has something more in mind, something more fundamental: a debate about public investment and sustainable growth.

Needless to say, I'm one of the cyclists. I agree that health care, energy, and income inequality (which I see as the result of a broken education system, but that's another issue) are long-term structural issues that need to be addressed, but I don't see these as the causes of the recession.

So what is the key point of disagreement here? In my view, it is the section I put in bold above. Dr. Reich is a Keynesian -- he believes that aggregate demand drives the engine of economic growth. I believe that investment is the principal driver of growth. In Dr. Reich's view, we need to shift money to those with lower income because they spend more and save less. In my view, we should encourage people to consume less and save and invest more -- invest in their health, in their education, and in their savings accounts. Savings and investment, in both human and physical capital, are the path to prosperity, not increased consumption.

There are a few other minor points of disagreement between our narratives of the US economy over the past 40 years, but in my view this is the most fundamental.

Brad DeLong's Answer to Tim

2009-01-29T16:13:00.004-05:00

Brad Delong gives his take on The Most Important Ideological Question of 2008:

The current recession may turn into a small depression, and may push global living standards down by five percent for one or two or (we hope not) five years, but that does not erase the gulf between those of us in the globe's middle and upper classes and all human existence prior to the Industrial Revolution. We have reached the frontier of mass material comfort—where we have enough food that we are not painfully hungry, enough clothing that we are not shiveringly cold, enough shelter that we are not distressingly wet, even enough entertainment that we are not bored. We—at least those lucky enough to be in the global middle and upper classes who still cluster around the North Atlantic—have lots and lots of stuff. Our machines and factories have given us the power to get more and more stuff by getting more and more stuff—a self-perpetuating cycle of consumption.

Am I still a Republican?

2009-01-28T19:19:00.003-05:00

Today the stimulus bill passed the House by a vote of 244 - 188.

Democrats voted in favor of the bill 244 - 11, while Republicans voted against the bill 0 - 177.

On strictly economic grounds, I rate the bill a B. Were I in charge, I probably would have preferred a somewhat smaller bill, though I understand the desire for erring on the larger side. I also would have cut out most of the spending portion that won't hit the economy until FY 2011, and replaced it with more direct cash transfers, such as payroll tax cuts. However, most of the spending hits in the first three years, and the non-stimulus portions are a relatively small part of the total bill.

Were I a member of the House, I think that I most likely would have voted in favor of the bill. Of course, there are political considerations in how one votes. Were I a Republican congressman, the House leadership probably would have put substantial pressure on me to vote with the party. I also might have been offended that my voice had not been heard in the process, or may have been trying to make a statement of some sort on the portions of the bill I don't like. Nevertheless, I still think I would have voted "aye."

This raises a rather uncomfortable question: Does this make me a Democrat?

Time for a Little Diplomacy?

2009-01-27T20:48:00.003-05:00

Paul Krugman says he's not a diplomat. That's a bit of an understatement, but he's not alone. The last week or so has seen a high level of non-diplomatic exchanges between economists. At issue is the huge stimulus package proposed by our new president, which will shortly be before Congress.

When the stimulus was first proposed, Joe Biden was quoted as saying that there were no serious economists who disagreed with the idea of the stimulus package. Greg Mankiw quickly corrected him, pointing to a long list of prominent economists who have expressed doubt in the package, including himself:

Greg Mankiw:

As I have documented on this blog in recent weeks, skeptics about a spending stimulus include quite a few well-known economists....If any university managed to hire all of them, it would immediately have a top ranked economics department.

Among the most prominent, besides Mankiw himself:

Robert Barro:

But, in terms of fiscal-stimulus proposals, it would be unfortunate if the best Team Obama can offer is an unvarnished version of Keynes's 1936 "General Theory of Employment, Interest and Money." The financial crisis and possible depression do not invalidate everything we have learned about macroeconomics since 1936.

Eugene Fama:

Government bailouts and stimulus plans seem attractive when there are idle resources....Unfortunately, bailouts and stimulus plans...do not add to current resources in use. They just move resources from one use to another. And bailouts and stimulus plans only enhance future incomes when the activities they favor are more productive than the activities they displace.

John Cochrane:

In sum, there is a plausible diagnosis and a logically consistent argument under which fiscal stimulus could help: We are experiencing a strong portfolio and precautionary demand for government debt, along with a credit crunch. People want to hold less private debt and they want to save, and they want to hold Treasuries, money, or government-guaranteed debt. However, this demand can be satisfied in far greater quantity, much more quickly, much more reversibly, and without the danger of a fiscal collapse and inflation down the road, if the Fed and Treasury were simply to expand their operations of issuing treasury debt and money in exchange for high-quality private debt and especially new securitized debt.

From there, the debate has become quite undiplomatic, and not terribly productive. To whit:

Paul Krugman:

First Eugene Fama, now John Cochrane, have made the claim that debt-financed government spending necessarily crowds out an equal amount of private spending, even if the economy is depressed....What’s so mind-boggling about this is that it commits one of the most basic fallacies in economics....So how is it possible that distinguished professors believe otherwise?

The answer, I think, is that we’re living in a Dark Age of macroeconomics. Remember, what defined the Dark Ages wasn’t the fact that they were primitive — the Bronze Age was primitive, too. What made the Dark Ages dark was the fact that so much knowledge had been lost, that so much known to the Greeks and Romans had been forgotten by the barbarian kingdoms that followed.

And that’s what seems to have happened to macroeconomics in much of the economics profession. The knowledge that S=I doesn’t imply the Treasury view — the general understanding that macroeconomics is more than supply and demand plus the quantity equation — somehow got lost in much of the profession. I’m tempted to go on and say something about being overrun by barbarians in the grip of an obscurantist faith, but I guess I won’t. Oh wait, I guess I just did.

Paul Krugman:

As the debate over President Obama’s economic stimulus plan gets under way, one thing is certain: many of the plan’s opponents aren’t arguing in good faith. Conservatives really, really don’t want to see a second New Deal, and they certainly don’t want to see government activism vindicated. So they are reaching for any stick they can find with which to beat proposals for increased government spending....It’s true that the normal response to recessions is interest-rate cuts from the Fed, not government spending. And that might be the best option right now, if it were available. But it isn’t, because we’re in a situation not seen since the 1930s: the interest rates the Fed controls are already effectively at zero.

That’s why we’re talking about large-scale fiscal stimulus: it’s what’s left in the policy arsenal now that the Fed has shot its bolt. Anyone who cites old arguments against fiscal stimulus without mentioning that either doesn’t know much about the
subject — and therefore has no business weighing in on the debate — or is being deliberately obtuse.

Brad DeLong:

Oh boy. John Cochrane does not know something that David Hume did--that the velocity of monetary circulation is an economic variable rather than a technological constant....

You sometimes see this mistake in freshmen students in Economics 1, students who do not fully understand either the circular flow of economic activity or what a credit economy is. They think--like Cochrane--that the flow of spending must be constant unless somebody "prints money" because, you see, you need "money" in order to buy things....

Cochrane's mistake--an elementary, freshman mistake--is because he has not thought enough about how a credit economy works....

Milton Friedman knew this. Irving Fisher knew this. Simon Newcomb knew this. David Hume knew this. John Cochrane does not know this: does not know that the velocity of
circulation is an economic variable rather than a technological constant.

Brad DeLong:

I am watching Eugene Fama, Robert Lucas, and now John Cochrane stagger around assuring everyone that fiscal policy cannot boost employment and production--no matter how high the unemployment rate and how much unused capacity there is -- as a matter of "just accounting" that "does not need a complex argument about 'crowding out'"....And I can barely believe my eyes....

Milton Friedman disagreed with James Tobin about the relative effectiveness of monetary and fiscal policy in "normal" times. I agree with Friedman (and disagree with Tobin) about the relative effectiveness of monetary and fiscal policy in "normal" times. But Friedman thought Tobin was worth debating--Friedman did not have the stupids to claim that Tobin was completely wrong as a matter of "just accounting."

Paul Krugman:

Economists who have spent their entire careers on equilibrium business cycle theory are now discovering, in effect, that they invested their savings with Bernie Madoff.

Hey, I never claimed to be a diplomat.

Paul Krugman:

There are certainly legitimate arguments against spending-based fiscal stimulus....

What’s been disturbing, however, is the parade of first-rate economists making totally non-serious arguments against fiscal expansion. You’ve got John Taylor
arguing for permanent tax cuts as a response to temporary shocks, apparently oblivious to the logical problems. You’ve got John Cochrane going all Andrew-Mellon-liquidationist on us. You’ve got Eugene Fama reinventing the long-discredited Treasury View. You’ve got Gary Becker apparently unaware that monetary policy has hit the zero lower bound. And you’ve got Greg Mankiw
— well, I don’t know what Greg actually believes, he just seems to be approvingly linking to anyone opposed to stimulus, regardless of the quality of their argument.

Needless to say, everyone I’ve mentioned is politically conservative. That’s their right: economists are citizens too. But it’s hard to avoid the conclusion that all of them have decided on political grounds that they don’t want a spending-based fiscal stimulus — and that these political considerations have led them to drop their usual quality-control standards when it comes to economic analysis.

Has there been any comparable outbreak of mass bad economics from good liberal economists? I can’t think of one, although maybe that’s my own politics showing. In any case, what’s happening now is pretty disturbing.

Mark Thoma [Usually a more moderate voice]:

Quoting an email [from Paul Krugman], economists who "have spent their entire careers on equilibrium business cycle theory are now discovering that, in effect, they invested their savings with Bernie Madoff." I think that's right, and as they come to this realization, we can expect these economists to flail about defending the indefensible, they will be quite vicious at times, and in their panic to defend the work they have spent their lives on, they may not be very careful about the arguments they make. I don't know if the defenders of the classical faith have come to this realization yet....

Much of this debate is unproductive, since it revolves around the legitimacy of the models under consideration. Most of the arguments quoted above amount to a restatement of the authors basic position, followed by an assertion that their model is clearly correct, and that their opponent's model is clearly incorrect or discredited. The conservative voices tend to argue that their opponents are simply Keynesians who haven't adapted their arguments to account for the literature of the past 30 years. The liberals reply that the last 30 years have been a dark ages of sorts, that their opponents' arguments are little more than reiterations of the "treasury view" discredited by the Great Depression, and it's time to bring back Keynes.

Lost amidst this furor has been the modicum of legitimate debate. These are occasions where one side or the other has laid out a model that does not imply by it's very nature that they are right and that the other side is wrong, but that allow for a framework for productive debate over the values of the model parameters. A few examples:

Menzie Chinn

Kevin Murphy

Brad DeLong's response to Kevin Murphy

Arnold Kling

I personally think that this approach is vastly preferable to the puerile arguments over modl superiority found elsewhere. Simply put, both sides need to create a model mutually acceptable to the other, rather than claiming that their highly respected PhD-wielding foes are violating basic accounting identities or making econ 101 mistakes. Politicians make mistakes when they take accounting identities and econ 101 too seriously. PhD economists should have a respectful debate over the parameter values of commonly accepted models.

There are two kinds of economics. There's the economics that is decided by ideology, and there is the economics that is decided by econometrics. We as a profession should be concerning ourselves with the latter.

****** EDIT *******

Arnold Kling responds to the "Dark Age" comment, and I think he's striking the right tone. Well worth the read. Over the past few weeks, I have been most impressed with Kling and Greg Mankiw's approach to the debate, at least among the econ bloggers I read regularly.

Liberals, Conservatives, and Prices

2009-01-26T15:39:00.002-05:00

I believe that many of the differences between liberal and conservative economists is their beliefs about how people respond to prices. Conservative economists believe that people are price sensitive, while liberal economists believe that people aren't very sensitive. More specifically, conservative economists believe that changes in price prompt large and rapid changes in behavior, while liberal economists believe that behavioral responses to price changes are slow and small.

Strong Keynesians demonstrate this quite nicely. Consider a recent op-ed by noted Keynesian economist and Clinton Labor Secretary Robert Reich. He writes glowingly of unions:

"Go back about 50 years, when America's middle class was expanding and the economy was soaring. Paychecks were big enough to allow us to buy all the goods and services we produced. It was a virtuous circle. Good pay meant more purchases, and more purchases meant more jobs.

At the center of this virtuous circle were unions. In 1955, more than a third of working Americans belonged to one. Unions gave them the bargaining leverage they needed to get the paychecks that kept the economy going. So many Americans were unionized that wage agreements spilled over to nonunionized workplaces as well. Employers knew they had to match union wages to compete for workers and to recruit the best ones."

At the core of this argument is a Keynesian model in which the driver of growth is demand. Keynesian economics argues that when aggregate demand increases, companies expand their operations to produce enough to meet the demand. As they do so, they hire more workers and employ more capital, which causes aggregate demand to increase even more.

In other words, Keynesians look at the GDP accounts equation:

(1) Y = C + I + G + X - M

And say, "Okay, so if we want to maximize Y we obviously need to increase C, I, G, and X, and decrease M." Thus Keynesian will seek to increase spending by government and individuals in order to prompt more investment, and they will argue for encouraging exports and discouraging imports.

Of course, (1) is an accounting relationship, not an economic model. As such it completely ignores prices. Any attempt to raise one of the components of GDP will affect all the others through prices. For instance, if wages increase, the cost of labor increases. If money is transferred from the rich to the poor to take advantage of the poor's lower savings rate, or if government raises taxes to increase spending and finances it by debt or taxes on the rich, savings will fall. When savings fall, the cost of capital increases.

Conservative economists argue that when the costs of production fall, firms will respond by expanding production, and when the costs increase they will cut back on production. Thus Reich's "cure" would do nothing but increase costs and decrease long-run growth. But if prices have little effect on incentives, and if the effect is slow in coming, then these policies would work.

This applies to just about every dispute between conservatives and liberals. If firms don't change their employment level much in response to wage increases, than minimum wages laws will not decrease the supply of jobs by very much, but will benefit workers by a lot. If workers are not very responsive to wages, than a decrease in after-tax pay (via increasing taxes) won't affect their incentive to work very much. There, we now have explained the conservative-liberal divide on tax policy and price controls.

In fact, when you get right down to it, nearly every policy difference can be explained in this manner. Conservatives don't like big government, because government must be paid for by taxes or debt, which lower wages and increase labor and capital costs. Not a problem if you believe that firms and workers aren't very responsive to these costs, but a big problem if you believe they are. Inelastic supply and demand curves mean small deadweight loss from government interventions, while elastic curves means large deadweight loss.

Conservatives argue that markets regulate themselves well, but that's just another way to say that they're responsive to prices. In the market model, prices are information signals, so saying that markets process information efficiently is the equivalent of saying that they respond to prices efficiently. By contrast, liberals believe that markets are really just a mishmash of irrational behavior and rent-seeking. As such, prices rarely convey useful information.

One way I like to think of this is that prices are waves through the medium of the economy. Liberals believe that the underlying structure of that economy is so complex that the signals become hopelessly muddled, and prices end up with such a high noise/information ratio that they do not regulate economic activity effectively. Conservatives believe that prices move through a relatively frictionless medium and convey good information to parties, thereby leading to coordination and adaptation to new information.

What's particularly nice about this formulation is that it allows for testable predictions. For instance, what's the noise/information ratio in equity and currency markets? Does it move with measures of fundamentals, or does it often rapidly depart? Are price elasticities in demand models generally high and signed correctly? Are partial adjustment terms in demand models generally high or low?

These are empirical questions, but unfortunately economics is a very difficult discipline to get good empirical results. Demand models are difficult because prices and quantities are endogenous, and instrumenting is more art than science. Structural demand models, such as those including partial adjustment terms, are likely to fit well due to temporal consistency, but that does not mean that they in fact capture the appropriate factors. Disaggregated data are hard to come by and rarely accurate and complete enough to perform the sort of econometrics we would like to perform.

This idea is sometimes expressed in a "People are rational vs. people are irrational" framework. This is a valuable way to think about it, but it's a bit more wishy-washy and less solid than talking about price responsiveness. People have tried to gauge rationality by neuro-economic experiments, etc. but this doesn't really answer the question of whether groups of people behave in ways that are little different from how groups of rational agents would behave. In other words, you only need rationality at the margins.

Personally, I take the view that people are rational within familiar frameworks, but that unfamiliar frameworks can bring out their irrationality. That is, people are optimized for behaving in familiar environments, but if the environment changes they will retain rules of thumb and habits that are no longer appropriate to the changed environment. The economy is constantly facing small shocks that prices accurately process and result in bear optimal market responses. Occasionally, however, large shocks hit that the market is not prepared to process.

Some of these shocks aren't really that big a deal. A shock in the form of a housing bubble bursting, such as we're experiencing now, is pretty small in the big scheme of things. You can argue about how we should respond in the short-run, but it will make only a few years difference in the end (that is, the difference between an optimal response and a sub-optimal response might be the difference between reaching year 2020 development in 2020 or in 2023), but it won't sink the system.

Other shocks are potentially more dangerous. For instance, will natural resources run out in a slow controlled manner, in which markets can respond optimally? Or will we reach a "tipping point" in which we'll suddenly realize we're out of oil?

Response to Tim F (part 1 of 2)

2009-01-25T16:18:00.005-05:00

In a comment on my response to his original post on The Most Important Ideological Questions of 2008, Tim F writes:

I cannot understand the disintegration of spiritual and material wellbeing that you spell out here. I believe the bible paints a picture in which all creation falls together and all creation is re-made together, so I don't see how, from a theological perspective, the conditions of material wellbeing could become disentangled from a general condition of fallen-ness in the world and start experiencing dramatic increases. I cannot figure out how to get your apparently reformed eschatology (capitalizing on the depravity of man in the present age) to square with this sunny humanistic infinite growth model of global economic value.

I am sure you have felt the need to understand the disjunction between the progress of material and spiritual wellbeing--what is your explanation?

I would like to start by saying that this is indeed an issue with which I have wrestled. When I was a child, I thought that the world's problems were technical, and so decided to become a scientist. Later I came to see the problems as social, and so resolved to become a social scientist, eventually settling on Economics. Now I see the world's problems as spiritual, arising from sin.

So where does that leave us? The only cure for sin is Jesus, so it follows that the only true cure for sin's consequences -- including all manner of economic brokenness, inequality, injustice, poverty, and environmental degradation -- is Jesus.

How do we reconcile this theological belief with the observation that material welfare has increased dramatically over the past 200 years? One approach is to reexamine the observation, and argue that the human condition has not materially improved since 1800. The second option is to reexamine our understanding of this theology in light of the evidence.

There can be reasoned debate about whether the last 200 years have represented true progress. I'm personally quite confident that they have, but that's for my next post. For now, let's assume that there has been such progress. Is this observation inconsistent with Biblical teachings about sin and salvation?

Our axioms are that (1) all life's problems arise from sin, and (2) the only cure for sin is Jesus. Does it then follow that we cannot do anything about life's problems other than seeking after Jesus in a personal sense, and leading others to do the same? I would argue that we can do more. Material problems can be addressed even if their source is not cured, in the same way that the symptoms of a disease can be treated even if the disease itself is not cured. Death will still come if the true cure is not found, but the process can be made a little less painful.

So when I speak of a distinction between material and spiritual welfare, I speak of the difference between solving the problem of sin, and seeking to fix some of the negative material consequences of sin. I don't think that it's unbiblical to suggest that, over time, we could become better at fixing some of these consequences.

God gave us minds to reason with and hands to work his Earth. While our relationship with the Earth has been broken by the curse, we still retain our nature, which is in the image of the divine. God's original directive to be fruitful is still in effect, and from my reading of Genesis it seems that God anticipated a growth in material prosperity over time, as Adam and Eve and their descendants created and built. I believe that this directive persists, albeit distorted by sin, and that we are seeing it in action now. Material wealth cannot save us, and shall turn to ash in our hands on judgment day. Yet there is true beauty at its root, and it can be redeemed if we use it to bless others and store up treasure in heaven.

God sometimes speaks through non-Christians; heck, sometimes he speaks through the mouths of donkeys! God sends the rain on the wicked as well as the just, and their crops grow. I even believe that God derives glory from the accomplishments of human minds even when those minds refuses to give glory to him. He created the mind, and all that is good and creative in it derives from him.

I do not believe that the growth of the past two centuries has been unambiguously good. Like all human endeavors it is full of sin, there has been evil in it. Yet I do not think that it is unbiblical to suggest that the material state of man has improved, nor that sin has entirely co-opted the Genesis growth model. This growth may still come, shrouded in sin like all human actions, yet its blessings still deriving from God.

The Hierarchy of Effective Stimulus

2009-01-23T11:49:00.003-05:00

There are basically three desiderata for a stimulus:

1. Timely: The sooner it hits the economy the better
2. Targeted: The more it preferentially targets unemployed resources and limits crowding out, the better.
3. Effective: Ideally, the spending would be for things we want or need, rather than pointless make-work.

Here are my thoughts on the various components of prospective stimulus plans:

What: Direct aid to state governments to make up budget shortfalls
Conclusion: Very good
Reasoning: This aid is timely because it will employ resources right now. It is focused, because it employs resources that would otherwise be cut back, meaning that there is zero crowding out. Finally, it is effective, because these programs were in place pre-recession, and so passed cost-benefit analysis in the past.

What: Direct cash transfers to the private sector via tax cuts and lump-sum rebates.
Conclusion: Good
Reasoning: This aid is pretty timely, although there are non-zero administrative costs in getting the money out the door. It can be focused if you aim it to exploit the same labor and credit market rigidities that perpetuate recessions in the first place, and if it's mis-targeted people will just buy bonds with it leaving us no worse off. Finally, it's effective, because it puts money in the hands of people who know best how to spend it in ways that will benefit them, whereas government spending may not be so efficient.

What: Limited infrastructure spending
Conclusion: Okay
Reasoning: Timeliness is a big concern with infrastructure, but there are projects at the margin that can be put in place quickly. It is pretty targeted, because the recession has hit the construction sector hard, meaning that there will be little or no crowding out at the margin. Limited spending is likely to be reasonably effective, since there are probably decent projects that were on the agenda and can be moved up. Whether it's more effective than tax cuts probably has to do with whether you believe we needed more infrastructure spending pre-recession, though there's little to lose at the margin due to reduced costs of employing these resources.

What: Large-scale infrastructure spending
Conclusion: Bad
Reasoning: Timeliness is now a huge concern, since there are only so many projects being planned right now. It will be less targeted than marginal spending, since such a large spending increase will likely begin to crowd out private construction pull resources away from other sectors of the economy. Effectiveness is also a concern, since a lot of new projects that weren't in the works would be planned hastily, and less emphasis would be placed on the benefits rather than getting the projects moving.

What: Major new government programs and restructuring, such as health care reform or "green" jobs.
Conclusion: Very bad
Reasoning: Simply put, this is not a stimulus, and treating it as a stimulus could be disastrous. It would not be timely, because it represents a large-scale shift in resources and would have to be carefully planned and considered. It would not be focused, since many of the jobs would be created in industries that are already expanding or at least not contracting, and may result in lots of crowding out, particularly over the long-run, as government takes over management of large swathes of our health care and energy sectors. Finally, in order to be effective, such industrial reorganization must be carefully and thoughtfully implemented with long-run goals in mind, which is the precise opposite of a stimulus.

If it were up to me, I'd draw the line after tax cuts, but I'm fine with some infrastructure spending too. However, I'm very concerned with too much more than that, since there is a high risk of poorly-directed ineffective government programs that arrive too late to do anything but crowd out private investment, or else programs that fundamentally affect the economy in the long-run, but were poorly and hastily designed in order to be included as part of a "stimulus" bill that can be ram-rodded through Congress.

Better and better or worse and worse?

2009-01-21T19:11:00.004-05:00

My friend Tim F. asks:

Is the world getting better and better or is the world getting worse
and worse?

This question is, he believes, "The Most Important Ideological Questions of
2008...your answer to this very basic question will determine a great deal of your politics and your religion. It will certainly determine your response to the many technological developments in which we are awash."

I agree that this question is very important. Moreover, I agree that
the assumptions implicitly contained within this formulation underly much disagreement in public and private discourse. While I'm not sure that it's a better delineator of political ideology than the question of whether our lives are more shaped by our environments or our choices, I do agree that it is a very useful categorical tool.

This is also an issue that is very close to my heart. As economist, I seek to understand society and to structure it to improve the material welfare of all. Has such improvement occurred in any consistent way throughout history? Is it now occurring? Can it occur in the future?

In addressing this question, the first task is to define "better" and "worse". I think that it's valuable to consider two different definitions. The first is the material definition -- a state of the world could be materially better if a certain basket of measurable indicators are better. For instance, if people are living longer and healthier lives, reporting more happiness in polls, enjoying more education and more leisure time, spending less time and energy fighting and struggling, and more time and energy on creating and enjoying art and other luxury goods rather than bare necessities. The second is the spiritual definition -- if people are finding more meaning in their lives, developing a closer relationship with God, and finding more peace and joy in interpersonal relationships.

I believe that both of these sorts of "goodness" can be found in life, and that both are contained within the Biblical notion of "Shalom." I believe that we as Christians should seek to better the world in both respects, in both large and small ways.

Now to answer the question of whether the world is getting better or worse -- I believe that the Bible teaches that there is no trend to speak of in aggregate spiritual welfare. The world is fallen, and the best we can do is try to nurture what goodness there is in the individuals we interact with daily. Perhaps as the church grows and God pursues his redemptive purposes, overall spiritual wellness may increase marginally. However, the end of spiritual poverty shall not come about until the return of Christ, and until then the best we can do is perform the role of salt and light -- i.e. a preservative and revealer.

But what of material wellbeing? Here we have something that can be measured. The result of such measurements show that there was limited material progress from antiquity until the onset of the Industrial Revolution. However, since around 1800 we have seen rapid increases in aggregate indicators of material prosperity, beginning in Western Europe, spreading to European offshoots, and finally beginning to spread, however unevenly, throughout the entire world. There are now few parts of the world that have not seen some increase in material wellbeing since the beginning of this growth explosion, although it has been uneven and its sustainability is in doubt.

So that's my take. Spiritual welfare is pretty much constant, and material welfare was pretty much constant until recently, when it increased dramatically albeit unevenly. However, I might be a bit biased by the fact that my profession is devoted to studying this economic growth, and that my career goals are to devote myself to ensuring its spread and sustainment.

WTF?

2009-01-12T20:42:00.004-05:00

This is percentage change in average real hourly earnings, per the BLS:

Wait, am I seeing things? Did real wages increase by 22% in November? Did they increase by 13% in October?

I'm assuming that these numbers are annualized, but still. That's ridiculous. The last time we had an increase of 13% or more was January 1972, when it increase 14%, which was the largest of all time. The current number is provisional, but seriously...WTF!

This is on the verge of convincing me that Casey Mulligan is right.

EDIT:
Sorry, this is "Change in Real Earnings", see the historical data on the right sidebar in the link above (click on the little green dinosaur). Change in average hourly earnings is a healthy 5%, but nothing compared to the change in real earnings.

EDIT 2:
Ah, stupid me. The numbers are given in 1982 dollars, not in percentage terms. So +0.22 means that average real earnings increased by $0.22. It's still the largest in history, though.

Left-Wing Lies and the E-P Ratio

2009-01-10T09:54:00.014-05:00

I'm here to call out a few major left-wing economists for something that I consider to be quite frankly deceitful. As much as I like to give people the benefit of the doubt on things, I have trouble believing these people don't know exactly what they're doing.

First I'll talk about the unemployment rate. The UR is one of the best statistics for measuring exactly where we are in the business cycle. Elevated unemployment indicates higher frictional forces, which suggest coordination failure in the labor market, which is a major indication of recession.

Here's what the unemployment rate has looked like since 1948:

The unemployment rate is calculated as the number of people who are unemployed and actively searching for a job divided by the number of people who are working plus those people. In other words, if we call people who have jobs plus those actively searching the "labor force", the unemployment rate is the fraction of the labor force that is unemployed.

Now you can see just by eyeballing it that the unemployment rate tends to follow the business cycle. Each peak in unemployment rate corresponds to a recession, and each trough corresponds to a boom. Furthermore, the unemployment rate seems to move within approximately the same range. The peak of the 2001 recession was lower than most other recessions, and the following trough was deeper than most troughs, suggesting that the 2000's weren't that bad from a cyclic perspective.

Of course, that doesn't sit well with the Left, since Bush was president during the 2000's. However, it turns out that there's another measure of employment that tells just the story the Left wants to hear. I'm talking about the employment-population ratio. This statistic is calculated by taking the total number of people who are employed, and dividing by the total number of people. For instance, here is the E-P ratio for everyone over the age of 16:

You can see why this appeals to the Left. It shows that times were best under Clinton, and that the "Bush boom" was worse than the late '90s boom. It plays into their narrative that the economy peaked for "regular people" during the late '90s, and just hasn't gotten back there since.

Furthermore, several prominent left-wing economists have been pushing the E-P ratio as a better measure of the business cycle than the unemployment rate. Just so you don't think I'm setting up a straw man, here are some choice quotations:

Paul Krugman:

Employment-population ratio

I think the picture speaks for itself.

Update: OK, maybe it doesn’t. What it shows is that the “Bush Boom” was really fairly pitiful compared with the boom under Clinton, both in magnitude and in duration.

[note that he only shows 1993 - 2007 so that there's no historical perspective. But that's another class of deceitfulness]

Matthew Yglesias:

As you'll see, the ratio dipped quite a bit at the beginning of the new millenium as the stock bubble burst, and then stayed substantially lower than it had been during the 1990s boom throughout Bush's much-touted recovery and now has started declining again without ever reaching the old peak or even the late-1970s peak.

Angry Bear:

Several economists have suggested that the ratio of employment to the civilian adult population (let EP = E/P) is a better measure of the strength of the labor market. EP was 64.4% when Bush took office but fell to 62.2% by August 2003. For many months, EP remained virtually constant even as the Employment Survey was showing moderately strong employment growth. Some have suggested that employment growth has barely kept pace with population growth.... I shall also suggest that this measure is a better measure than the unemployment rate....

Brad DeLong:

Add them all up, and discover that since 2001 our two main labor market indicators--the unemployment rate and the payroll employment count--have been giving us contradictory readings. Given where the unemployment rate is now, we would under normal conditions expect payroll employment to be 3.5 million higher than it is. Given where payroll employment is now, we would expect the unemployment rate to be 1.7% higher than it is now. The unemployment number is the number we would expect to have with a 134.8 million job economy (but we actually have a 131.3 million job economy).

Brad DeLong also recently referred to the E-P ratio as his "favorite chart."

I've talked about the stupidity of the Left's employment fetishism in the past, but here's a quick review: we had close to 100% employment back in the year 1000 AD, because everyone engaged in back-breaking toil from birth until death just to scrape out a living. Number of jobs isn't an indication of anything, and in fact decreasing the number of jobs may be a good thing, if it means that people get a chance to retire or go to school instead. The purpose of long-run macroeconomic policy should be to maximize labor productivity, meaning the value of production per person, because this is how you get a wealthy society with a high quality of life. Productivity is increased by investment in human capital (health and education) and physical capital, including public goods.

While employment statistics can be useful measures of short-run frictions in the labor market, they should not be a goal of long-run policy. The unemployment rate strictly dominates the E-P ratio for this purpose.

Just to drive home this point, let's take another look at what's been driving the E-P ratio lately. Here it is broken out by gender:

So the increasing trend in the E-P ratio over the past 60 years have been the number of women entering the (measured) workforce. Of course, these women hadn't been sitting at home twiddling their thumbs for the past 1,000 years. Instead, there used to be a large section of the economy that took place at the home. Things like cooking, cleaning, and child-rearing, as well as many forms of home production, used to be much more labor intensive and time-consuming than they are today. Let's take a closer look at what's been happening to the male E-P ratio:

So there has been a steady decreasing trend, as more men over the age of 16 spent time in school or lived longer and longer after retirement.

Here's the women's E-P ratio:

This has been increasing for a while, but has recently leveled off. The departure of the unemployment rate and the E-P ratio has been caused by this leveling off. This caused some rather strange numbers to show up during 2003 - 2006, in which the number of jobs was not increasing as much as the fall in the unemployment rate would suggest. However, as this chart makes clear, the reason was that the demographic trends during the 2000's were fundamentally different from the past. The problem wasn't men leaving the labor force at higher rates than in the past, but rather that women's participation rates had topped out at a number a little shy of men's.

Now some might argue that there still is a gap between men and women which needs to close, but I'm not so sure. Once you factor in the decisions by many women to take extended time off work to raise their children, plus the fact that women retire younger and live longer after retirement, I'm not at all sure that we haven't reached a new long-run equilibrium. It may be that the women's E-P ratio will soon begin to follow the men's downward trend. This isn't a problem, and we shouldn't determine policy based upon it. Instead, we should try to maximize long-run productivity growth, and look to the unemployment rate as the cyclical indicator of choice, even if it has the cheek to not conform with our partisan prejudices.

Free Healthcare in Ghana

2009-01-06T17:37:00.004-05:00

A new study finds no significant improvements in health outcomes of providing free health care in rural Ghana:

This trial in rural Ghana found that children in households randomised to free healthcare used formal healthcare more and informal healthcare less than a control group. This
utilisation did not translate into any change in anaemia (the primary outcome), mortality, or other health outcomes measured. An observational group who had paid to selfenrol into the same scheme were wealthier, healthier, and used both formal and informal healthcare more than those
randomised to it at baseline and subsequently. A number of studies in both developed and developing countries have investigated the impact of lowering direct financial barriers
to health care on utilisation, but this is the first randomised trial to investigate the impact of providing free health care on health outcomes. It used malaria-associated health outcomes
in children as the indicator of health impact as it is the most important cause of serious childhood mortality and morbidity in the area. The failure to find any demonstrable health
benefit from the change in utilisation following free health care was demonstrated even for those living within 5 km of a health care facility (so with limited physical barriers to
access). This lack of any effect, including on secondary outcomes such as Hb for which the study had good power, challenges the assumption that where introducing free health care leads to changes in utilisation, it can safely be assumed to translate into health benefits. Given the potential size of resources involved in providing free health care that could be diverted from other priorities on the basis of that assumption, this finding is potentially important for policymakers.

I would not conclude from this study that medical treatment had no effect, since the control group also accessed care. Rather, it suggests that there was no significant marginal benefit of seeking additional care beyond that which would be sought in the absence of free access. Furthermore, the fact that the self-selected exogenously wealthier group performed better on all health measures indicates that the operative factors are wealth and its correlates, which likely include education.

Overall, I see this study as favoring an approach to development focused on increasing the wealth and education of the poor, rather than decreasing the cost of services. Since free care represents a transfer of wealth, there is an income effect that increases the welfare of the recipients; however, a strictly better result could be achieved by simply giving the money to the recipients directly. Finally, all participants would likely benefit from a hospital in their district, indicating that health infrastructure is important as well.

EDIT:

The study also states:

There was a
slight increase of 5.2%, 4.2%, and 7.2% in insecticide-treated net use among the control, intervention, and self-enrolled arms, respectively over the study period, but these increases were not significantly different.

Although these results were not statistically significant, they accord with economic theory and suggest one facet of a possible mechanism that may dilute free health care. These numbers suggest that those with access to free medical care were less likely to use mosquito nets to protect themselves from malaria. Some economists have suggested that this sort of decision-making, called moral hazard, is behind a lot of the apparently counterintuitive results regarding the low marginal benefits of health care access. This theory suggests that giving someone free access to health care encourages risky behaviors that can leave health outcomes unchanged, or even worse.

For instance, suppose that the mosquito net costs the equivalent of a week's income, while malarial treatment costs the equivalent of a month's income. The family with free health care may decide to forego the nets, and use the money to buy food or a goat. The family with no free health care may pay for the net since contracting malaria is more costly for them than it is for the other family.

Why Casey Mulligan is Wrong

2009-01-02T14:53:00.032-05:00

Casey Mulligan caused a bit of a stir with a Christmas Eve Economix post, in which he wrote:

Unlike in the severe recessions of the 1930s and early 1980s, productivity has been rising. Through the third quarter of 2008, productivity had risen six consecutive quarters, with an increase of 1.9 percent over the past three, or 0.7 percent above the trend for the previous 12 quarters.

Because productivity has been rising — almost as much as the Douglas formula predicts — the decreased employment is explained more by reductions in the supply of labor (the willingness of people to work) and less by the demand for labor (the number of workers that employers need to hire).

Central to this analysis are the implications of aggregate supply and demand in the labor market.

If you draw neat supply and demand curves, as you might find in a textbook, you'll conclude that productivity rises when the supply curve shifts inward and falls when the demand curve shifts inward. However, the real aggregate curves need not behave so nicely.

Consider aggregate supply and demand curves* S and D1 that intersect at an equilibrium level of employment q*. Now suppose that the demand curve shifts inward, resulting in a new curve D2, s.t. D2(q) < D1(q) for all q in N. The demand and supply curves intersect at a new equilibrium q, with q < q*.

Here is a picture of what that might look like:

GDP is the area under the demand curve. Thus, before the contraction GDP was x + y + z, and after the contraction it becomes z. Furthermore, employment falls by dq, winding up at q (so q* = q + dq).

Productivity is defined as GDP per worker, or GDP/q. Since both GDP and q fall, it is unclear whether productivity increases or decreases. In particular, productivity falls if (x + y + z)/(q + dq) > z / q

Multiplying each side by (q + dq) * q we get:
(x + y + z) * q > z * (q + dq)

Subtracting z * q from each side, we get:
(x + y) * q > z * dq

or:
(x + y) / dq > z / q

This gives us an intuitive picture of productivity:

At a glance, we can see that productivity was higher before the contraction in demand. However, this is not always so. In cases in which x is small, productivity can increase quite substantially. Observe:

One might reasonably object that the first example looks much nicer and more similar to the pictures one finds in economics textbooks, while the second is contrived to produce the expected result. However, we're talking about a highly aggregated demand curve, and there is no reason to expect it to be nice and neat. In fact, one might reasonably expect it to be quite lumpy in places.

Moreover, there is no reason to expect the demand curve to contract uniformly. The contraction in demand for labor is not due to a productivity shock; rather, it is due to an unanticipated drop in consumer spending. This means that each worker produces as much as he does before, but firms are anticipating slack demand. Thus marginal workers appear very expensive to them, since they are producing goods that will not be bought. However, the more productive workers higher up the scale are producing just as much as before, and firms anticipate being able to sell what they are producing. Therefore, demand for workers at higher wage levels may not diminish by very much.

In other words, there has been an across-the-board wealth reduction, which has lead to an across-the-board contraction in demand for consumer goods and services. However, labor demand need not contract uniformly. It may contract primarily at the margin.

Why does this differ from past recessions? My theory has to do with the rising education premium, which has increased differentiation among workers. When there was a lower education premium, the aggregate labor demand curve sloped down primarily due to declining returns to physical capital. Now, however, it slopes downwards because of that and because of human capital stocks unique to the workers. Therefore we expect a steeper slope, as well as more general lumpiness of the demand curve. In fact, we might expect to see a two-tiered slope as portrayed above, corresponding to the gap between college graduates and non-graduates. ***

However, I do anticipate productivity falling in the near future. The reason is that people who have been laid off from contracting industries will soon be seeking employment in other industries in which they have less knowledge and experience. These "underemployed" will work for lower wages, and this shift will constitute destruction of part of the stock of human capital, which will lower productivity.

Note that I do not think that Mulligan is 100% incorrect. Certainly it is possible that the supply curve has shifted as well, and he has raised an important question about why productivity is increasing. It will be interesting to track productivity in the next few months, along with other measures.

* The labor demand curve is defined as the locus of points (w,N) at which the wage (w) is equal to the expected marginal product of the Nth unit of labor. The labor supply curve is likewise the locus of points (w,N) at which the wage w is equal to the marginal expected disutility of the Nth unit of labor.

** An interesting note: If we were to add a line for wages, which would lie between the supply and demand curves, we could divide GDP into three parts. The area below the demand curve and above the wage curve would be surplus to the employers -- i.e. that would be paid to the owners of capital in direct profit. The area below the wage curve is collectively labor income. However, the supply curve divides this into two sections. The area above the supply curve is the workers' premium, but the area below the supply curve corresponds to the portion of GDP that is not "surplus" to society, since that is what is necessary to compensate workers for the disutility of their work. This has significance, because this area makes it theoretically possible to increase GDP while decreasing the total surplus. In other words, it's quite possible to increase GDP without increasing welfare. I shall discuss this in more detail in a future post.

*** This has also lead to an increase in wage differentiation. We shouldn't expect a flat wage, curve here, but rather one that slopes quite a bit. Also, note that, since we're to the right of the big lump, we should see most of the job losses among the lower tier, i.e. people without college degrees. This may not be perfectly correct in this recession since it hit the finance industry pretty hard. However, many of those people may now displace lower-education workers in underemployed positions. One problem with increasing the education premium is an increase in economic inequality and, more pertinent to this analysis, vulnerability of the less educated.

Keynesian Resurgence?

2008-12-30T17:45:00.008-05:00

Wikipedia helpfully informs us that the recent economic history has lead to a Keynesian resurgence:

The Financial crisis of 2007-2008 cast doubt on the free market consensus. In March, free market guru Martin Wolf, chief economics commentator at the Financial Times, announced the death of the dream of global free-market capitalism, and quoted Josef Ackermann, chief executive of Deutsche Bank, as saying "I no longer believe in the market's self-healing power." Shortly afterward influential Economist Robert Shiller began advocating robust government intervention to tackle the financial crises, specifically citing Keynes.

My views are a bit different. I think that the important development has been the collapse of the Neoclassical Synthesis.

One feature of the synthesis was a tacit division of labor. The Keynesians, who were principally macroeconomists, concerned themselves with short-run macroeconomic models using aggregated data, while the neoclassicals, principally microeconomists, concerned themselves with disaggregated models and the long-run. This allowed the Keynesians to avoid the critique of the rational expectations literature, and allowed the Neoclassicals to concern themselves with real factors, and thereby ignore monetary and fiscal policy.

The present crisis, however, has once more focused the attention of economists on cyclical factors. This has naturally lead to increased attention on Keynesian economists. Moreover, this has caused the synthesis to come apart at the seams, as the crisis has lead Neoclassical economists to offer their opinion of the present downturn, and has lead Keynesian economists to critique the economic policies that lead to the current crisis.

This collapse of the division of labor has made it clear that models based on Neoclassical and Keynesian theory prescribe quite different courses of action, and that attempts to develop micro-foundations for Keynesian ideas have failed to satisfy the Neoclassical objections. This leaves economists with two different and contradictory theories, which presently widely divergent prescriptions for economic policy.

This unfortunate predicament reveals the intellectual shallowness of the original synthesis, as well as the pressing need for a unified economic theory to resolve the debate once and for all. The truce has ended, and now we are left to wonder whether one side shall triumph over the other, or whether a new theory -- a true synthesis -- shall emerge to win the day.

What is Truth?

2008-12-29T22:30:00.004-05:00

Quid est veritas?* quoth Pilate, and his sentiment haunts humanity to this day. Philosophers have been of many opinions, but I thought I would take a moment to share a few of my own thoughts on the subject.

Consider the concept of proof in mathematics. Mathematical proofs are often a hodgepodge of plain English, formal English, and symbolic representations. The mathematician wishes to satisfy his reader that a statement is "true", although its truth is not immediately apparent. In order to do this, the mathematician convinces his reader that the statement is equivalent to a statement whose truth is apparent, usually by a series of logical deductions.

Yet every step of this process relies upon the ability of the reader to perceive the "truth" of each statement. There is no way to compel this affirmation; rather, the mathematician must rely on his audience's knowledge, intelligence, and honesty.

At this point we might conclude that truth is no more than affirmation -- a collective fiction if you will -- in which case our strength of belief corresponds to our belief that intelligent people may perceive cases in which truth or falseness is apparent, and be consistently accurate in these perceptions.

This by itself doesn't bring us to a complete notion of truth. In order to accord with our intuitions we need truth to correspond to reality. Unfortunately, there is no way human beings can be the judge of this correspondence, so it is not immediately clear that this helps us at all.

However, there is an objective judge of claims of correspondence to reality. This judge is reality itself. If we assert that gravity does not exist, and then behave as though gravity does not exist, then we shall soon learn of our error or else perish in our ignorance. Reality favors those whose notions of truth best correspond to itself.

By definition, truth must triumph over falsehood. If A is true and B is false, then acting as though A were true must be better than acting as though B were true. Perhaps this explains Jesus' words to Pilate: omnis qui est ex veritate audit meam vocem.**

* What is truth
** All who are of the truth listen to my voice

Is Material Wealth a Good Thing?

2008-12-14T20:40:00.003-05:00

I've touched on this subject a few times, but it bears repeating. I believe that material wealth is good.

This is an important point. I am undertaking a career as an economist, in which I intend to devote myself to finding answers to my field's most important problem, that of understanding economic growth.*

I will confess plainly that I set out with great trepidation. One of my greatest fears has been that I would devote my life to a cause only to find, once my life is spent, that I had been of more harm than help to those I sought to serve. After all, "what does it profit a man to gain the whole world but forfeit his soul"?

There is a common meme in our society that material wealth is in some sense immoral. While this idea is sometimes justified biblically, I utterly reject it. I refuse to romanticize poverty or the poor, but seek a better material state for all mankind. While I do not undertake this mission with any hope of success, I see it as fulfilling my essential calling to do what I am.

I should be clear that the virtue I see in material wealth does not imply that the wealthy are virtuous. While certain virtues lend themselves to material wealth, less savory attributes do so as well, and there are other virtues that lead to poverty. Personally, while I have no particular plans to live in material want, I have no particular ambitions or expectations of personal wealth. My ambitions are larger and, I hope and pray, wholly submitted to God, such that they are now one and the same as God's ambitions for humanity.

I am not a transhumanist hoping to fashion a better future through human effort. Nor do I hope to tap into some mystical spirit of transcendence, the magic of the human mind. No, I am motivated by a simple belief -- that material prosperity is preferable to material want, and that providing for all of mankind's material needs, in an equitable and sustainable manner, is a noble objective. Or, to put it more succinctly, I believe that life is valuable and should be valued. I am pro-life.

If I am wrong, and if material wealth is itself corrupting and harmful, then all my professional efforts will be harmful to the world's spiritual health. But if this be so, I am content in the knowledge that my hope rests not in my own efforts, and that God will work it for good.

-----------------
* I consider the problem of economic growth to encompass both issues of distribution and of sustainable production. While these are often treated separately, I believe them to be merely two parts of the whole.

Charles Wheelan's Greatest Frustration

2008-12-13T12:45:00.006-05:00

It seems that Charles Wheelan agrees with me on the most important problem in the field of economics. He writes,

My Greatest Economics Frustration: There's an entire sub-discipline of economics devoted to the field of economic development. And there are huge organizations staffed with economists who work on global development issues, such as the World Bank.

Then there's Africa, which is shockingly poor and, with some exceptions, drifting sideways or even getting poorer. Economists can do a reasonable job of explaining how places like Japan developed, or how India is getting richer now. And they can document all the things that contribute to wealth: education, infrastructure, good governance, empowerment of women, and so on.

But the discipline is frustratingly bad at taking a particular place at a particular time -- whether it's the Detroit region or the nation of Chad -- and offering a prescriptive set of specific steps for growth and prosperity.

I agree 100%. So much good can come of solving the problem of economic growth.

Edit:
It seems that at the same URL Charles Wheelan lists the greatest economic challenge of our age as "Reconciling the tradeoff between a decent safety net and the bad incentives that it creates." I am disappointed. The possible utility gains of a better safety net for the millions of relatively well-off citizens of the developed world traded off against the possibility of faster national growth in these same countries is dwarfed by the possible utility gains of elevating billions of people out of poverty in the third world. The latter poses substantial unsolved economic problems that have received relatively little attention from economists. The former has received lots of attention, is fairly well understood, and much of the remaining disagreements are of a moral rather than technical nature.

Second Edit:
I should also mention that the tradeoff between growth and the safety-net is not even the most important problem facing the first world. I consider that to be the question of sustainable development, which encompasses the tradeoffs between economic growth, environmental destruction, and the limits of natural resources.

The Sentiment of Mr. Blagojevich

2008-12-09T20:54:00.004-05:00

Yesterday I happened upon a despicable quote by Mr. Blagojevich, the governor of Illinois. You may have heard of him since, but the statement to which I refer was made Monday, December 8, 2008.

It was in reference to the closing of Republic Doors and Windows, a factory in Chicago, IL. The plant is in financial distress and has just informed its workers that they will all be laid off without proper benefits. The workers are staging a sit-in in protest, and their plight has attracted the attention of Democratic politicians throughout Illinois and the country.

Interestingly, the wrath of the politicians has largely focused on Bank of America, not Republic. Bank of America provided the money for the venture, and the company blames the bank for failing to provide the financing necessary to keep the factory open.

This brings us to Mr. Blagojevich, who said Monday:

"How come it’s OK for some of these banks who have messed things up in a significant, substantial way that it’s trickled down to impact the lives of ordinary, average Illinoisans and ordinary, average Americans who every day work?"

"Somehow, these workers always end up on the bottom. It’s wrong...And if the taxpayers who do all the work in our country have already been asked to bail out these big banks, then we expect those big banks to bail out businesses like this … to keep these workers working."

Incidentally, CNN has since edited the article to remove the rather grammatically unflattering quotation, although it still can be foundhere. Nevertheless, it is the content that I want to critique.

What I find striking is Mr. Blagojevich's assumption that, because the bank was bailed out using public funds, it should now expend those same funds to keep an unprofitable business afloat. This is completely nonsensical. The government is perfectly capable of giving out money for political reasons. The whole point of banks is that they invest in profitable enterprises. To compel a bank to invest in an unprofitable venture for political reasons is to nationalize the bank.

Providing unemployment insurance and enforcing laws regarding severance pay is the job of government, not private financiers. As Bank of America properly argues, they had provided funding and have now stopped, as is their prerogative and as Republic knew would happen. Whether or not Republic defaults on its obligations to its employees is not Bank of America's concern, because the fulfillment of the obligations of a creditor is not the responsibility of the lender. Lenders do not lend because the creditor needs; they lend because there is profit to be made.

The sentiment expressed by Mr. Blagojevich reminds me of the crony politicians and power brokers in Ayn Rand's Atlas Shrugged, and the quote evoked in me the same deep disgust I experienced when I first read that book. Over the next few days, Mr. Blagojevich will be roundly criticized by many of the same politicians who were so quick to stand by him Monday. In my eyes, the sentiment expressed by Mr. Blagojevich is indistinguishable from his crimes, and any politician who expressed a similar sentiment Monday and now condemns him is a hypocrite.

A Profound Statement

2008-12-07T17:59:00.017-05:00

Have you ever been talking with someone when they suddenly say something incredibly profound? I had such a moment a few months ago. In passing, a coworker said something along the lines of, "Economic theory allows us to draw inferences from economic data. That's what separates econometricians from statisticians."

This is very important, because economists can't apply the scientific method as it is traditionally used in the natural sciences. The scientific method is usually stated as:

(1) Theory --> Hypothesis
(2) Hypothesis --> Experiment
(3) Experiment --> Data
(4) Data --> Verification/Rejection of Hypothesis
(5) Verification/Rejection of Hypothesis --> Modification of Theory

Statisticians have a lot to say about (4), which is typically called "inference." Inference is not that difficult given a well-designed experiment. Unfortunately, economic "experiments" can't really be designed. Therefore, the 'economic' method is a bit different. It runs something like this:

(1) Theory --> Hypothesis
(4) Data --> Verification/Rejection of Hypothesis
(5) Verification/Rejection of Hypothesis --> Modification of Theory

Steps (2) and (3) are dropped, because there is (usually) no way to conduct an experiment to directly address the hypothesis. Instead, the economist collects data, or more usually finds data collected by others, that are related to the area of interest. The emphasis is not on experimental design, but rather on the process of inference, which usually takes up a lot of time and effort. Regressions are run, relationships are charted, and a good dataset may provide the raw material for several publishable papers.

I have nothing against econometrics -- in fact, I think that econometricians have made a lot of important contributions to the field of statistical inference; however, I do believe that this process is more prone to error than the scientific process, and that economists need to be aware of these shortcomings. Using theory to interpret data is a tricky business, since we want to use the data to infer something about the validity of the theory. The economist must be wary of the process of inference collapsing into a tautology.

I believe that this problem is one of the chief causes of the persistence of major disagreements among economists. Often, data are presented that, when interpreted through the lens of a particular theory, corroborate that theory. However, if data cannot discriminate between theories it is inferentially useless. Yet this is often obscured by the implicit rather than explicit use of the theory in the interpretation.

What is the solution? I don't have a complete answer, but as a first approximation economists should put more effort into applying theory before looking at the data. Ideally, economists should be able to agree on how results will weigh on the respective likelihood of competing theories independent of the process of interpreting the results.

To put it more plainly, I believe that economists should pay more attention to 'inferential design.' This should properly be done before the data are examined or even collected. Hopefully treating inferential design explicitly, and separate from actual inference, will lead to the resolution of outstanding problems in economics in a manner satisfactory to all practitioners.

Advice from China

2008-12-07T10:09:00.002-05:00

As reported in the Financial Times, our business partners across the Pacific are full of advice:
Zhou Xiaochuan, governor of the Chinese central bank, urged the US to rebalance its economy. “Over-consumption and a high reliance on credit is the cause of the US financial crisis,” he said. “As the largest and most important economy in the world, the US should take the initiative to adjust its policies, raise its savings ratio appropriately and reduce its trade and fiscal deficits.”

Let's look at the national balance of payments equation:

S - I = X - M (Savings - Investment = eXports - iMports)

At the moment, the US has a big trade deficit, meaning that X - M < 0, and that therefore S - I < 0. In other words, we're importing a lot more than we're exporting, and we're financing this via foreign capital flows, which allows domestic investment to exceed domestic savings.

What's driving this state of affairs? While there are other factors at play, one of the largest is the policy of the Chinese government of purchasing US public debt in order to devalue the Renminbi relative to the dollar. By dumping cheap Renminbis in the US, the Chinese encourage elevated imports of Chinese goods. Therefore both our trade deficit and our fiscal deficit are the consequence of the policies of the Chinese central bank, of which Mr. Xiaochuan is governor.

If we are going to "rebalance" our economy, we need to recognize that imbalance is impossible in the aggregate, and that the imbalance of any individual nation is only possible if there is a corresponding imbalance in another nation. It takes two to be imbalanced, and it takes two to rebalance.

For years China has been ignoring US requests to allow the Renminbi to appreciate to an equilibrium level against the dollar. Early indications are that the current economic downturn will significantly hurt China, possibly even more than the US. This is in large part due to the Chinese economy's heavy reliance on exports to the US, which makes it particularly vulnerable to downturns in US demand. While this is unfortunate from a global welfare perspective, it is the natural consequence of an imbalanced economic policy that seeks to maintain unsustainable levels of export-driven growth via currency devaluation. In the end, the Chinese government has nobody but itself to blame.

Lump Sum Forgiveness?

2008-12-06T14:08:00.002-05:00

Ben Bernanke wants the government to do more to stop foreclosures.

U. Chicago's Casey Mulligan warns that mortgage forgiveness would create adverse incentives.

And I'm left wondering where the Coase Theorem failed.

Note: The post's title refers to Mulligan's statement that any mortgage forgiveness should be undertaken "without regard for borrower circumstances." If forgiveness has the same incentive effects as an income tax, he's essentially advocating that it be undertaken lump-sum. A more academic version of Mulligan's thoughts on the subject is available here.

Industrial Restructuring

2008-12-05T19:27:00.010-05:00

In my post If Brad DeLong Doesn't Know..., I wrote:
By contrast [to economists and politicians], each person who actually works in the industry has expert knowledge of some component of the industry that nobody else possesses, nor could hope to match. It is these people who will be doing the actual work of restructuring.

Does this mean that there's no room for government intervention in economic processes? Absolutely not. Market processes can fail, and certain government interventions can, under certain conditions, improve the state of the economy.

To illustrate these failures, consider the economy as a collection of individuals each exercising the resources at their disposal to better satisfy their utility functions. Due to the shape of human utility functions, there tends to be clumping as individuals find utility in certain interactions with other individuals. We refer to these interactions as 'market exchanges'. Some of these clumps we might refer to as 'markets.' Larger ones we might call 'industries', and even larger ones 'national economies'. And finally we refer to the whole shebang as the 'global economy.'

Now one useful thing about markets is that they are theoretically efficient, meaning that, under certain assumptions, individuals maximizing their own utilities will reach a state that is collectively 'optimal', according to a particular definition of optimal.

Unfortunately, individual space is very large, and deciding where to go next is a very difficult thing. Each individual has a search function, whereby they search possible states and decide what to do next. But their possible moves are constrained by their location, and the search functions needn't be particularly optimal.

Given this model, we can wind up the economy and let it run and it will eventually settle in some nice optimal equilibrium. However, this equilibrium needn't be universally optimal, since starting point matters. Also, this equilibrium needn't even be locally optimal if the market assumptions fail to hold.

Thus, we can broadly speak of two sorts of market failures. The first takes place when we settle into a locally optimal equilibrium, but we could do better somewhere else. The second occurs when, even though we are not at a local optimum, individual actions aren't moving us towards it.

The solution to both of these comes through the framework of collective institutions. People who are thinking on a global level can identify nearby optima and try to effect one another's optimization processes in hopes of moving there, though we might need to go downhill a bit first.

What this means is that, although the actual work of moving from one location to another is carried out on the individual level, there is room for institutions.

Now, I've written a lot on this blog about the second kind of market failure -- ones caused by the failure of the assumptions of efficient markets. These are the sort that are most broadly studied by economists, and we have done a good job of categorizing the more common errors, and created many sorts of institutional solutions.

The first kind of market failure, however, is less commonly studied. It's dealt with a bit in the endogenous growth literature, when people speak of industries' growth paths and whatnot; but it certainly hasn't reached the level of academic attention as the more common variety of market failure.

So Brad DeLong's trepidation is well-founded, because when we speak about industrial restructuring, we're talking about trying to move to a very different optimum, and that will take some serious adjustments to the institutional framework.

However, I think it's very important that we keep in mind that the process whereby we move to a better location will still be driven by individual action. The total optimization power held within the millions of people working within these industries far out-strips that mustered by economists and politicians. The emphasis should be on channeling this power, rather than ignoring it.

Institutions cannot pick an industry up and drop it in a new location; they can only set it on a new path, heading towards a distant goal, and hope that we will eventually achieve it.

If Brad DeLong Doesn't Know...

2008-12-05T08:47:00.008-05:00

Brad DeLong writes,
How Are We Going to Manage to Do All This

The Obama administration is going to be rebuilding and reconstructing five major sectors of the American t. It has no choice--there is no other option. It has to remake:

* Autos
* Housing finance
* High finance
* Energy
* And the big one—health care

On what principles and through what procedures is this extraordinary exercise in structural economic reform policy going to be accomplished? I get how to do the macroeconomics of Obama administration economic policy. I don’t get how to do the structural side…

If Brad DeLong doesn't know, then who does? And if nobody knows, how is the Obama administration going to "remake" five major industries?

Economists know a lot about the principles of economics, which allows us to say very interesting and useful things about a lot of industries from a generalized intellectual framework. Moreover, there are many Economists who specialize in applying this framework to certain contexts and to particular industries; but they still do not have direct experience in the industry (with a few exceptions). While they can say a lot about restructuring in general terms, they can speak with authority only on a general level.

Politicians, on the other hand, know how their constituents think about economic issues, and they (hopefully) consult with economists and businessmen to get a grasp on how an industry operates. While this should give them a general idea of how an industry functions, it certainly is not enough to "restructure" it all by themselves.

By contrast, each person who actually works in the industry has expert knowledge of some component of the industry that nobody else possesses, nor could hope to match. It is these people who will be doing the actual work of restructuring.

Was the Depression Avoidable?

2008-12-03T09:29:00.004-05:00

Here I'm differentiating between the recession that has been going on since (officially) December, and the sharp downturn that began in September. To differentiate, I will be referring to the latter as a depression.

The recession, much like the dot-com recession of 2001, started out with an asset bubble. In this case, the assets in question were real estate and houses. The origins of the bubble have been variously ascribed to zoning ordinances, low interest rates, lax regulation, the securitization of mortgages, greed, and government efforts to encourage home-ownership among ethnic minorities. More likely a combination of these started housing going up, and then bubble mania took hold. Once the bubble got past a certain size, a painful correction was inevitable.

Initially, the recession manifested itself in the contracting construction and real-estate markets. From there, it began to spread to other areas, but seemed fairly contained. Then it was discovered that we had been improperly pricing the risk of securitized mortgages, that the toxic stuff was all over the place, and that we weren't really able to figure out exactly where it was. This began to erode confidence in sectors of the financial industry, and that whole house of cards began to unwind. Soon it became clear that fairly large-scale realignment was going to take place in the financial industry.

Thus far, this story has been a story of a recession -- a painful adjustment that is healthy and natural. However, at this point the story takes a decidedly darker turn. The problem, and it was a depression-grade problem, was deleveraging.

Simply put, a lot of people had underestimated the risk of certain assets, and therefore had become overleveraged. When they began reassessing this risk, they began to deleverage, which means selling assets and building up capital reserves. Naturally, per Gresham's law, they were inclined to sell their lemons. Both the deleveraging and the lemons began to cause problems.

With everyone trying to sell assets and few trying to buy, the capital stocks of financial institutions, generally valued at market prices, were eroded. Due to the lemon problem, the assets on the market were valued at a discount, further lowering the value of financial institutions' holdings. This meant that banks that had reserve requirements, or financial institutions that were looking to hold more reserves, had to do even more to acquire those reserves, leading to a vicious cycle of deleveraging.

This was a depression-grade problem, but it wasn't an obviously depression-grade problem (I'm defining recessions as things that market forces correct, and depressions as things that market forces worsen). It was possible that this deleveraging would bottom out, and that the financial system could get on with realignment. However this was not to be.

The proximal cause of the depression was the failure of Lehman Brothers. You see, even though a lot of assets had been seriously downgraded in peoples' eyes, there still was a lot of confidence in these institutions because of the implicit government guarantee that they were too big to fail. By violating that guarantee, government massively devalued a whole class of assets. This delivered a powerful shock to the financial system, as suddenly a whole lot of assets became a whole lot riskier. This kicked the vicious cycle into high gear, leading to a full-blown credit crunch. By the time the tepid bailout plan made it through Congress (on its second try), we were well into a depression.

Would the depression have taken place had the government bailed out Lehman's? Frankly, I don't think we'll ever know. However, with the benefit of hindsight, I believe this depression could have been averted had the following steps been taken:

1. Bailout Lehman's.
2. Make explicit the government guarantee to bailout other major financial institutions and make good on their obligations.
3. Put through Congress a bill placing serious money at the Treasury's disposal to make good on this guarantee.
4. Began negotiating to purchase common stock (perhaps jointly with private parties) in troubled institutions.

I believe these steps would have averted the worst. But they weren't taken, and now we're in a depression. So now what? Well, that's a sticky question. Depression economics has been widely studied but is poorly understood, because depressions aren't that common and it's hard to draw good empirical conclusions from such limited data (the general problem with macro). I'll outline the debate a bit in my next post, and then give my own opinion.

Time for An Experiment?

2008-12-02T21:20:00.003-05:00

Is it a time to experiment with economic policy? Dani Rodrik thinks so:
FDR was an experimenter who understood the need for institutional innovation during uncertain times. Here is what he said in May 1932:

"The country needs and, unless I mistake its temper, the country demands bold, persistent experimentation. It is common sense to take a method and try it: If it fails, admit it frankly and try another. But above all, try something. The millions who are in want will not stand by silently forever while the things to satisfy their needs are within easy reach."

Repeat after me: "bold, persistent experimentation."

Russell Roberts thinks not:
Bernanke doesn't seem to realize (nor does Paulson) that constantly changing strategies reduces the desire on the part of investors and borrowers to take risk.

Afraid to look passive, policy makers risk making things much much worse. It is a good time right now to talk less and probably even to do less. But the political incentives all point in the opposite direction.

Roberts also cites a paper by Robert Higgs, which is well worth the read, although you should keep in mind that his thesis is not universally accepted among economists.

Personally, I think it's a bit of a stretch to talk about "regime uncertainty" in such a broad way. Uncertainty about property rights is not the same thing as uncertainty about monetary and fiscal policy. Higgs argues that the former discouraged private investment during the 1930s and prolonged the Great Depression, but I'm not sure that his argument has much bearing on experimentation with monetary and fiscal policy.

Tax Cuts or Infrastructure?

2008-11-26T10:41:00.003-05:00

Mark Thoma weighs in on whether tax cuts or infrastructure spending are a more appropriate stimulus.

The optimal level of long-run government infrastructure spending is a separate issue from how this spending should be implemented on a cyclical basis. I think that we can all agree that government has been underspending on infrastructure lately, and so the optimal level of spending is higher than it is now. A second question, however, is whether infrastructure spending is a more efficient stimulus.

I hope we can all agree that government spending comes at a cost, since the money has to come from either tax cuts or sale of government assets, which crowds out private consumption/investment. This is clearly worth it at least some of the time, since public goods are under-provided by the market. However, is the cost of government spending pro-cyclical or counter-cyclical?

The argument that the costs are pro-cyclical goes something like, "During recessions there are slack resources, so some share of the resources activated for infrastructure projects were previously unemployed. Therefore there's less crowding out, so we should spend on infrastructure off-cycle."

The counter argument goes something like, "Recessions are times of productive re-alignment and restructuring, so the long-run costs of crowding out investment could be even higher than during boom times, when we're seeing over-investment, so crowding out might actually be a good thing."

This hinges in part on your view of recessions in general, and this recession in particular: If recessions are times are natural and healthy periods of restructuring, then government infrastructure should be neutral or pro-cyclical. If recessions are coordination failures, then spending should be counter-cyclical.

Personally, I'm pretty much in the middle on this. I would advise that we increase infrastructure spending to what we consider to be the long-run optimum, and not raise tax rates. So this would take us to where we would be if we had had responsible on-cycle policy, and then decided to cut taxes in response to the recession.

Edit:
I should add:

When I say "tax cuts", I'm mostly envisioning increasing unemployment insurance, plus maybe stimulus checks preferentially targeted towards lower-income folks. This way you circumvent Ricardian equivalence and get people to actually spend. We also need to fight deflation big-time, but that's another issue.

Non-Economics

2008-11-14T10:57:00.003-05:00

Eliezer Yudkowsky has an interesting post over at Overcoming Bias:

There's a tradeoff between the strength of the assumptions you make, and how fast you learn. If you make stronger assumptions, you can learn faster when the environment matches those assumptions well, but you'll learn correspondingly more slowly if the environment matches those assumptions poorly.

....

...absence of solution implies an immense design space, and it requires a correspondingly immense amount of work by the programmers to reimpose order. Under the subsumption architecture, it's the programmer who decides to add an asynchronous local module which detects whether a robotic leg is blocked, and raises it higher. It's the programmer who has to make sure that this behavior plus other module behaviors all add up to an (ideologically correct) emergent intelligence. The lost structure is not replaced. You just get tossed into the Turing Tarpit, the space of all other possible programs.

On the other hand, CES creates order; it adds the structure of probability distributions and gradient optimization. This narrowing of the design space takes so much work off your hands that you can write a learning robot in 137 lines (at least if you happen to be Sebastian Thrun).

....

Added: Robin Hanson says, "Economists have to face this in spades. So many people say standard econ has failed and the solution is to do the opposite - non-equilibrium instead of equilibrium, non-selfish instead of selfish, non-individual instead of individual, etc." It seems that selling non-apples is a full-blown Standard Iconoclast Failure Mode.

I agree with Robin Hanson here. Neoclassical Economics is a framework that allows a high degree of power in understanding how the world (of human interactions) works. Although it's highly flexible in most normal applications, it's highly inflexible in a universal sense, which is the source of much of its power.

Pointing out the shortcoming of neoclassical assumptions has become a parlor game, but nobody has proposed an alternative with anything close to its power. Until they do, neoclassical Economics will rule the roost, and the various heterodox schools will be left out in the cold.

Incidentally, my personal parlor game is showing how neoclassical assumptions are actually very good. But that's another matter.

No Free Lunch!

2008-11-13T09:54:00.003-05:00

KNZN writes:

When an economy has slack resources, as the US economy – as well as the world economy – clearly does now and likely will even more in the immediate future, there is no aggregate welfare cost to using up those resources, so any benefit society receives is, in the aggregate, free. In an aggregate welfare sense, slack resources are a free lunch, and that lunch is wasted if nobody eats it.

This is incorrect.

When we talk about resources, we're talking about productive factors, which means labor and capital ("idle men and idle machines"). However, there are two possible causes for slack resources.

The first is coordination failure. This occurs when, for some reason, the Economy is failing to properly match labor to capital. In this case, there is a strong case for government intervention in order to overcome this failure.

However, it is also possible that slack resources exist because the Economy is going through a process of restructuring, during which contracting industries shed labor. This results in a painful transition period, as well as a destruction of capital (both physical and human) through disuse. However, the Economy is ultimately the better for it (paging Dr. Schumpeter).

Of course, in any Economic downturn it is likely that both causes come into play, since realignment leads to transitional coordination failures, and since hard Economic times accelerates realignment. Given that the current downturn is due in part to a credit crunch, there is a strong argument in favor of coordination failure as the dominant cause. Nevertheless, there is doubtless some much-needed restructuring going on, and certainly failing industries are hit hardest (I'm looking at you, GM).

So, in conclusion, government spending to mobilize slack resources is not costless, because it interferes with restructuring. It is valid to the extent that coordination failure is the cause, and invalid to the extent that Economic restructuring is the cause.

Of Bailouts and Stimulus

2008-11-11T18:46:00.004-05:00

A week ago, Barack Hussein Obama was elected the 44th president of the United States of America.

This is a heck of a time to become president, and it didn't take the press long to compare Mr. Obama to Franklin Delano Roosevelt.

Now what? It's pretty clear that we're heading into one of the biggest Economic Recessions since The Big One. The financial system is in dire straights, which suggests that a heavy dose of medicine may be needed to get things moving again. As the sickness spreads from the Developed to the Developing World, all eyes turn towards Washington, saying, "Physician, heal thyself."

It's an exciting time to be an Economist.

Two big issues are currently on the table. The first is the question of what sort of stimulus the Economy needs. The second is whether and how to bailout currently struggling industries. You can generalize these questions as, "Should the government inject capital into the Economic system, and if so how should it do so?"

This is a very complicated question. In fact, it's no exaggeration to say that this is one of the fundamental debates in Economics. Every cent of capital that is directed by government is a cent that is not directed by the private sector; it has to be, since the government cannot simply magic capital out of thin air. However, this is not a conclusive case against stimulus of all sorts, except in the eyes of a few hard-core Libertarians.

Thus far, the debate has raised more questions than it has answered. The reasons is that Economists have many rules of thumb and ideological views on how much capital to inject, but the bottom line is that we really don't know how the Economy works on this level. Thus the debate has at time degenerated into a debate over the causes and cures of the Great Depression, FDR's legacy, etc. etc. But the bottom line is that there is still a huge debate between the Keynesians and Monetarists. There is no neoclassical synthesis, merely a redrawing of old battle lines.

So what, dear reader, is the answer? Do we need a huge stimulus, or a small stimulus, or no stimulus at all? Do we bail out Michigan? Do we create a New Deal II? Do we send out checks in the mail and extend unemployment benefits?

At this juncture, my answer is simple: I don't know. My best guess is that we will take a politically-motivated scatter-shot approach that will benefit various interest groups, and that we will have a short and sharp recession followed by a lengthy recovery. This will be followed by a slew of papers by Economists explaining why events validated their personal theory, suggesting that we really didn't learn anything at all.

Born Again?

2008-11-05T09:41:00.002-05:00

I love reading exit polls. Very interesting stuff.

One split from CNN:

White Protestant Born Again: 73% for McCain
White Protestant Not Born Again: 54% for McCain
All Others: 31% for McCain

Politics aside, I find the idea of people self-identifying as "not born again" quite disturbing.

And not to keep you in suspense, the first category was 23% of voters, and the second was 19%.

Is it Time for a New Keynesian Economic Order?

2008-10-27T10:42:00.003-04:00

The Democratic critique of American economic policy over the past several decades has focused chiefly on increasing inequality. In the past, this has principally focused on the exclusion of the poor (disproportionately minorities) from access to basic goods and services, such as health care, housing, and education. As gains in productivity arising from technological advances and deregulation have gone disproportionately to the wealthiest members of society, Democrats have argued for increasing taxes on the rich so that government can provide these services.

Yet this tune has changed dramatically in recent years. While before the emphasis was on the exclusion of the poor from the wealth of the Pax Americana, this has shifted to the erosion of the middle class and the problems arising from deregulation. This is really a huge difference as an Economist.

You see, the old argument was more a question of philosophy than of Economics. There was a consensus among Economists that deregulation had allowed the private sector to drive growth, and that redistribution through high taxes came at the cost of growth. Conservatives argued that these costs were high, that inequality was actually lower than it seemed, and that there was no strong moral imperative to redistribute income; Liberals argued just the opposite. Nevertheless, the basics of this consensus were not seriously questioned among mainstream policy economists.

Yet the new critique is radically different. It is not a question of philosophy, but one of economics. Liberal Economists have begun to question the approach of private-sector-driven growth, calling for government to take a strong role in making investments in infrastructure and green technologies to "create jobs" and determine the growth path of the Economy. Likewise, redistributionist arguments have been justified on efficiency grounds as "bottom-up" growth, a radical departure from the idea of a tradeoff between growth and equality. Finally, deregulation has gone from being seen as one of the chief driver of growth, to public enemy #1.

These perceptions are widely shared in the press and among common Americans. They are fed by fear: Fear of jobs going to China and India, of illegal immigrants, of falling industrial output, of a weakening dollar and strengthening Euro. Fear of debt, of rising medical costs, climate change, and falling home values. Some of these fears are legitimate, and some solutions do require government intervention. Yet I believe that we are in prime position for a massive overreaction. There is a strong risk that America will take an isolationist route, and will fall back to Mercantilist and Keynesian policies, justified by the institutionalist critiques found in New Growth Theory.

Inequality and the Poverty Trap

2008-10-24T09:50:00.005-04:00

In my previous post, I wrote that poverty shouldn't exist, at least not in the presence of perfectly functioning markets. In fact, in the presence of perfectly functioning markets, a person's aggregate income is only a function of his innate ability and effort, plus interest payments on savings, which is really just earnings on physical capital. Thus any inequality is a function of individual choices and characteristics, as well as gifts between individuals.

Needless to say, we do not live in a world with perfectly functioning markets. However, we do live in a world with functioning markets in many goods and services, albeit imperfect and incomplete. Moreover, the level of imperfection in these markets varies widely among populations. Thus it might be a reasonable guess that poverty is somehow connected with the functionality of markets.

Unfortunately, the degree of completeness in a market is highly correlated with total quantity of economic transactions. This means that poverty drives away the very mechanism by which poverty is alleviated, leading to persistent poverty among certain populations. This effect is well-understood among development economists, and is referred to as the "poverty trap."

The poverty trap is usually thought of in terms of extreme poverty, i.e. "Extreme poverty does not allow markets to function properly." However, the same can be said of inequality. When a society includes gross levels of inequality, most of the transaction volume is attracted to the sphere of the wealthy. If markets were perfectly functioning, any opportunity for pure economic profit would be snapped up in short order; but in reality, large profit opportunities remain open for significant periods of time. Thus it isn't worthwhile to focus on capitalizing on small profit opportunities while the big ones exist. Given a high level of inequality, more profitable opportunities will always exist among the rich, and so the Poor will find themselves excluded from access to functioning markets.

To illustrated this, take the example of micro-finance. It turns out that making small loans to individuals and small organizations in developing countries can be profitable. In the theoretical world of perfectly functioning markets, these opportunities would be snapped up in short order. However, it just isn't worth it for Wall Street execs to tramp through the third world looking for small entrepreneurs who could use a loan while there are . Even though there's profit to be made, why bother when there's more profit to be made back on Wall Street?

Of course, micro-finance does exist, albeit it's a fairly recent innovation. It exists because there are ex-Wall Street execs who are willing to tramp through the third world looking for small entrepreneurs who could use a loan, albeit for non-monetary reasons. It also exists because of the efforts of some to scale up micro-finance opportunities to attract the big bucks.

So this allows us to rephrase questions about inequality and poverty as follows: How do we rectify inequalities in access to functioning markets?

Poverty Shouldn't Exist

2008-10-20T19:34:00.007-04:00

Poverty should not exist.

This isn't a moral sentiment; it is a mathematical fact. If you look at basic Economic models, you will find that in a perfectly functioning free-market economy, lifetime earnings are a function of three things: natural ability, a willingness to work, and luck. And a well-functioning insurance system will eliminate luck.

Of course, I'm not speaking of wealth on a global aggregate scale, but rather its distribution over a given region. If a country has limited capital and technology, then it will remain impoverished; however, if a country is rich, there shouldn't be poor people within that country. If a planet is rich, there shouldn't be a poor nation upon it.

The reason for this is simple -- in an Economic system with perfectly functioning markets, nobody is excluded from taking the steps necessary to climb out of poverty. In the presence of perfectly functioning credit and insurance markets, there is no luck left. If people are born in a poor country, or a poor city, they can borrow money against their future incomes in order to move to a better location. If people can't afford to go to school, they can borrow money against their future earnings for tuition. If people don't have the capital to start a business, they can borrow money against their business' future success, and take out an insurance policy to pay down their debt should they fail.

As you've probably guessed by now, we don't live in this world. We do not have perfectly functioning markets in goods, credit, or insurance. But it's nevertheless important to keep in mind that poverty is a market failure. It shouldn't exist.

What's Wrong With This Thinking?

2008-10-06T19:11:00.002-04:00

Ah, the compelling logic of Keynesianism. It keeps sneaking into anything the Media write about the Economy.

Consider a story that appeared today on CNN:

And Banerji said that the increasingly grim view of the economy will by itself lead to cutbacks in spending by both consumers and businesses. That in turn will result in greater job losses and more economic pain.

Can you spot why this is incorrect?

It certainly sounds reasonable. Afterall, if people and businesses spend less, there will be less aggregate demand. If there's less demand, businesses will buy less and cut wages and jobs. Then people will make less money, and the vicious cycle continues. This is the heart of the Keynesian view of Economic cycles.

So what's the problem with this? Simple. If people quite rationally decide to cut back on spending because they anticipate a decrease in their expected wealth in the near future, what happens to this money? Instead of being spent, it gets saved! Consumers will take their money out of their least favorite consumed products, and focus on buying what they like or need most, and all the rest will go into the banks. The banks, awash in extra capital as companies and consumers cut back, will cut lending rates in order to get the stuff moving again. Even as some established giants falter due to low demand for their goods, entrepreneurs will suddenly find themselves able to finance their ideas. So even as the least useful companies contract or go bankrupt, new companies arise to take their place.

As these new companies expand, they will look to hire new workers. Fortunately, lots of people will be laid off from the less useful companies, so wages will be depressed and the new companies will be able to hire employees at a bargain. Soon these employees will gain expertise in new useful fields, and new companies and industries will grow, supplanting the least useful of the old guard.

Schumpeter called this process "Creative Destruction." It is the process by which the Economy is self-regulating, and it underlies the idea that Recessions, though painful, are good for the Economy. Rather than viewing Recessions as terrible things in which a small shock can cause the whole thing to fall apart if government doesn't step in and pump up aggregate demand, Classical Economics views the Economy as a self-regulating system which forces things back to equilibrium.

However, there is one dire threat to the mechanism described above -- it assumes that people will save their money in banks, which will drive down lending rates. However, if the financial system itself is threatened by the Recession, like the AIDS virus targeting the human immune system, the mechanism by which equilibrium is restored may be unable to assert itself. If people don't have confidence in banks, they won't save their money, but will stash it under their mattresses (in reality, this means buying safe assets with low returns, like T-bills, or else buying commodities); worse yet is if the Recession hurts the banks themselves, such that they have to raise rates as things get worse (for instance, if there were lots of iffy Mortgage Loans taken out), we get the opposite effect. Suddenly the entrepreneurs have a harder time getting funding. If this process continues unchecked, the whole thing goes down in flames (see: 1931).

And that is the true danger. That is why I don't sweat much over rising unemployment and falling consumer spending, but I get terrified when banks start falling. There are a few government interventions that are reasonable in the former case -- mostly things designed to cushion the blow a little bit, and help to make the transition less painful. The latter case, on the other hand, is a code red, all hands on deck scenario.

Which is why I'm very, very afraid right now. Two weeks of non-liquidity have already taken us from a minor Recession to a very bad one. If things don't start working soon, more drastic government action may be necessary. And I'm not sure that I trust our government to regulate responsibly, particularly in an election year, given the amount of Economic disinformation and downright stupidity out there. They've already proven that they're not taking this doomsday scenario seriously with their failure to pass Bailout v1.0. And now that they've passed Bailout v1.1 (just like 1.0, but with more pork!), they seem more interested in assigning blame for the crisis than dealing with it. Which may simply be a function of where they possess expertise. Afterall, the democratic process is better at rewarding people who are good at passing on blame than people who understand Economics.

These are dangerous times. Please people, listen to your Economists!

The Most Important Problem in my Field

2008-10-06T14:21:00.002-04:00

Eliezer Yudkowsky:

Richard Hamming used to go around asking his fellow scientists two questions: "What are the important problems in your field?", and, "Why aren't you working on them?"

Often the important problems look Big, Scary, and Intimidating. They don't promise 10 publications a year, or promise any progress at all. You might not get any reward after working on them for a year, or five years, or ten years.

And not uncommonly, the most important problems in your field are impossible. That's why you don't see more philosophers working on, say, reductionist explanations of consciousness.

Trying to do the impossible is definitely not for everyone. Natural talent is only the ante to sit down at the table. The chips are the years of your life. If wagering those chips and losing seems like an unbearable possibility to you, then go do something else. Seriously. Because you can lose.

I'm not going to say anything like, "Everyone should do something impossible at least once in their lifetimes, because it teaches an important lesson." Most of the people all of the time, and all of the people most of the time, should be trying to do things that are possible.

Never give up? Don't be ridiculous. Doing the impossible should be reserved for very special occasions. Learning when to lose hope is an important skill in life.

But if there's something you can imagine that's even worse than wasting your life, if there's something you want that's more important than thirty chips, or if there are scarier things than a life of inconvenience, then you may have cause to attempt the impossible.

There has always been something in my heart stirred by the thought of taking on the impossible; something that seeks the glory of toil after a righteous cause, even if it is ultimately frustrated; something that is stirred by the myth of Sisyphus, and by Genesis 3:17-19.

So what is the Most Important Problem in Economics?

It is the simplest question of all: Why are some people rich, and others poor?

Of course, "people" can refer to individuals, but it can also refer to groups of people, whether defined by race, creed, or geographic location. Or even to the entire human race: Why are we richer today than 40 years ago (if indeed we are)?

To understand the depth of this problem, we have to understand that standard Economic models do not predict that a wealthy society would contain poor people; nor a wealthy country poor provinces; nor a wealthy world poor countries. So why is there poverty? And is there anything we can do to end it?

Housing Prices

2008-10-02T14:16:00.004-04:00

Should Congress do something to stabilize housing prices?

This is an interesting question. The argument for government intervention hinges upon a fairly fundamental question: do market prices reflect reality, or do they reflect something other than reality?

The big argument goes something like this: Housing prices fall until certain housing loans are under water, meaning that the houses are worth less than the total amount still owed. Consequently people default, either because it's not worth it for them to hang on to the home, or because they were counting on being able to refinance or resell. The banks seize the homes as collateral, and try to sell them to recoup as much of the loan as they can, eating the difference. The additional homes on the market drive the prices even lower, causing more defaults. This continues the vicious cycle.

In a functioning market, rising supply forces down prices, which causes producers to stop producing as much of the good. This eventually stabilizes price at the new equilibrium. However, in the story outlined above, falling prices are causing more houses to go to market, driving down prices below their optimal equilibrium. The proposed solution is for government to step in and renegotiate the loan contracts of people facing default, so that they can ride out the storm, causing prices to stabilize, and eventually rise again. With Uncle Sam stepping in with loose monetary policy, cheap loans from government institutions, and zoning regulations, housing prices would be on their way up again in no time.

However, I call b.s. on this story. There are four arguments for why government should not step in to restructure these loans:
1. This story depends on government being able to identify when a "reverse bubble" is in effect.
2. If restructuring were truly a valid option, banks wouldn't need government to tell them.
3. There really isn't any excuse for falling home prices to drive people to default at such high rates. Buyers will eventually be able to stabilize the prices on their own, and get cheap deals in the process.
4. The only case in which #3 does not hold is if people can't get credit to buy houses. So we're back to this being a liquidity problem, and the solution to that is a restructuring of the financial industry, not help to individual households.

Let's dig more deeply into each of these:

1. This story depends on government being able to identify when a "reverse bubble" is in effect. This is really at the core or most objections to government interference in the market. Market prices have a certain air of rightness about them, as they represent what a self-interested party is willing to pay. Government is not a self-interested party in the same way; instead, government is a constituent-interested party, with a limited understanding of Economics (derived from their constituents) and a strong incentive to appear to be doing something in the face of a crisis. Trusting government to decide at what price something should be valued is a recipe for disaster. It is the first step towards a command Economy, and a step that is disastrous in most other circumstances. A price control by any other name is still a price control.

2. If restructuring were truly a valid option, banks wouldn't need government to tell them. Why does a bank want to take a loss on a mortgage? If houses are currently undervalued, the home-owners have an incentive to find a way to hang onto them, and banks have a reason to not want to sell them now. So restructuring the deal should be entirely practical. Furthermore, this is already being observed in some instances. I see no reason that government intervention should accomplish anything that the free market cannot.

3. There really isn't any excuse for falling home prices to drive people to default at such high rates. Buyers will eventually be able to stabilize the prices on their own, and get cheap deals in the process. Given #2, we are forced to conclude that prices are being driven down because they're still overpriced. In this case, defaulting might make sense, but there's no reason to think that defaults would drive prices below fundamental levels. After all, homes are not only assets, they are also consumption goods. If someone bought a house, there's no reason that a fall in that house's value should make him unable to keep his mortgage unless he or she was either purchasing multiple houses as a speculator with the intention of reselling, or taking a loan he or she couldn't afford with the intention of refinancing later. There may be some people in this position, but it's not the case with all homeowners, so we shouldn't expect defaults to vary with housing prices in a linear fashion, so at some point this should bottom out. Also, since the homes are overpriced, having their value fall is good for many prospective homeowners, who ought to be able to buy up the homes and eventually stabilize the housing prices at the new equilibrium.

4. The only case in which #3 does not hold is if people can't get credit to buy houses. So we're back to this being a liquidity problem, and the solution to that is a restructuring of the financial industry, not help to individual households. Since the hero in our story of price stabilization is not the government but rather the home-buyer, the key to price stabilization is to make sure that he or she is able to access credit. For that, we need to keep liquidity in the financial system. By doing so, we allow houses to shift hands from people who shouldn't own them (speculators or those who can't afford them) to people who should (people who will buy them at reduced prices), with the prices stabilizing at equilibrium.

Presidential Leadership

2008-10-01T19:52:00.004-04:00

One striking thing about the failure of the Bailout has been the lack of leadership in our government. Voting for an essential but unpopular bill is basically a prisoner's dilemma scenario, as there is a tremendous temptation by individual Congress members to free ride on the votes of their peers, and defect for individual political advantage.

Although the Congressional leadership was generally in favor of the bill, there's no way that a bill this important should have any chance of being stopped. The following things needed to happen that did not happen:

1. Every major Administration official or advisor working in an area related to the Economy should have been on the record as saying that this Measure was critical.

2. Congressional leadership should have stepped up and made it clear in no uncertain terms that this measure was necessary, and that they expected 100% adherence, with substantial penalties for failure to comply. They should have made it clear to members of Congress that those who failed to vote for the bill would be blacklisted and not given positions on Committees, and would have restricted access to party funds in their reelection bids.

3. The presidential candidates should have played a bigger leadership role. At present, Congress is fractured and the President highly unpopular. McCain was 100% right to suspend his campaign, and Obama (and all the sycophants in Congress and the Press who agreed with him) were wrong. Both campaigns should have been suspended, and the debate should not have taken place. The candidates should been in Washington throwing every ounce of their political weight behind it, making speeches and twisting arms to get this thing through.

Although I think that McCain's decision to suspend his campaign was the right one in the circumstances, I'm not willing to give him too much credit. In the end he allowed himself to be swayed by the negative opinion of the press and the attacks of Obama. He should have stayed true to his decision and made this thing happen. His party is particularly lacking in leadership right now, as Republicans up for re-election seek to distance themselves from an unpopular president. But he failed, along with the rest of the political leadership.