Thursday, January 12, 2012

Alan Krueger Tries to Justify Income Redistribution

Here's a speech from yesterday by Alan Krueger, on income distribution. It's clear this is an attempt to provide some economics to support the ideas in a previous Obama speech. Maybe we should not give Krueger a hard time about his speech, as he is just out there doing his job. However, I think honesty and good economics are important, so let's see what Kreuger has to say.

It is by now well-known that there has been an increase in the dispersion of income in the United States, beginning perhaps as early as 1970. Krueger discusses this, but he also wants to make the case that there is less mobility across the income distribution than there once was, and less mobility in the United States than in other countries. Kreuger cites as evidence some work by Miles Corak at the University of Ottawa. Corak's "Great Gatsby curve" which shows a positive correlation between income inequality and immobility across levels of income within a country. Of course, this only establishes a correlation that exists in the data. Kreuger is making a policy speech. What we really care about in this instance is the effects of particular policies. To evaluate those, we need a serious structural model on which we can run experiments to evaluate alternative policies and compare their effects on economic welfare.

Krueger tells us about some of the causes of the increase in income inequality in the United States. This is pretty standard, though he has a funny way of assembling the evidence:
In the mid-1990s, I did a poll of a nonrandom group of professional economists attending a conference at the New York Fed. I asked them the extent to which various factors contributed to the rise in inequality.
Hopefully everyone understands why it is a bad idea to take a poll among economists to get a serious answer to any question. If I want to understand why income inequality has increased in the United States, I will read the relevant peer-reviewed published research, sift the arguments, and then draw some conclusions. Fortunately, in this case Krueger actually came up with an answer that is consistent with the received research. If Krueger had asked the same group of people to each write down their estimate of the government spending multiplier, I can assure you that he would get nonsense.

As is well-known, there are three key factors driving the increase in income dispersion. These are technical change, the scarcity of skilled workers, and import competition. Krueger, for political reasons, wants to attribute some blame to the Bush (W) tax cuts, but I think it is well-recognized that the effect of the change in the income tax schedule in this instance is relatively minor. He also talks about union membership and the minimum wage.

Here's where he starts to go off the rails:
Now, I could see why someone could support tax cuts for top income earners if they had materially benefited the U.S. economy, but the macro evidence is clear that the economy did not perform better after last decade’s tax cuts than it did after taxes were increased on top earners in the early 1990s. I already showed you evidence that income growth was stronger for lower and middle income families in the 1990s than it was in the last 40 years overall. This next chart shows that there was more job growth in start-ups in the 1990s than in the 2001-2007 period [Figure 11]. Across all businesses, job growth was much weaker in the 2000s than in the 1990s. So there is little empirical support for the claim that reducing the progressivity of the tax code has spurred income growth, business formation or job growth.
It is well-known as a theoretical proposition that the income tax has negative incentive effects. The key question, though, is how large those effects are. For more on this see this previous post of mine, particularly the part on Diamond/Saez toward the end. I think one can make a case that the incentive effects are large, particularly in the long run. When Krueger points to the fact that the Bush tax cuts were followed by poor economic performance, we know that's not serious evidence, as there were too many other things going on over that period.

In the next part of his speech, Krueger wants to tell us about the harmful effects of inequality. This starts off OK, by appealing to our sense of fairness. Maybe this income inequality is denying people opportunities? We could be seriously misallocating resources if high ability poor people are not being educated while low ability rich people are going to Harvard. But Krueger comes up with some pretty strange ideas as well. The first strange idea he attributes to Raghuram Rajan, which is that high income inequality encourages "families to borrow beyond their means." I haven't read Rajan's book, but Krueger could be mischaracterizing Rajan's ideas. My understanding is that, rightly or wrongly, what Rajan is arguing is that government attempts to redistribute income, working in part through Fannie Mae and Freddie Mac, contributed to the financial crisis. Krueger makes Rajan sound more like Robert Frank, which involves a very different set of ideas.

The second strange idea, attributed to Robert Reich, is that the increase in income dispersion is bad because it reduces "aggregate demand," since high-income people save more than low-income individuals do. To buy this idea, you have to think that we collectively make the wrong consumption/savings decision, and that redistributing income from rich to poor will move us toward a better national allocation of income between consumption and savings. Imagine a world with two people. A has 5 banana trees and B has 20 banana trees. A spends all his working time picking bananas and eating them. B spends half of her working time picking bananas, and the other half of her time planting new banana trees and tending to them. Krueger thinks the world in which A and B live would be better if someone took bananas away from B and gave them to A. I have no idea why he thinks that.

So what policies does Krueger have in mind? The first is health care, which of course is already done, and scheduled to be fully-implemented by 2014. There is a clear redistributive aspect to the Affordable Care Act. Those who stand to benefit from it are the poor, and it will be paid for disproportionately (because of progressive taxation) by the rich. I know this is controversial, but I don't think it should be. The Act won't do anything much to deliver health care more efficiently in the United States, but I don't have a problem with it. I'm from Canada. A second thing he pushes has to do with the American Jobs Act, which again we know about already.

Krueger then gets into some controversial territory - the financial industry and taxation. First, Krueger says "we must adequately regulate excess risk-taking and corrupt practices in financial markets." No one would argue with that statement. Who wants too much risk-taking or corruption in financial markets? However, what does that mean? How do we know excessive risk-taking when we see it? What do we do about it? Financial firms are of course very good at hiding "corrupt practices." How do we root those practices out? How do we tell useful financial innovation from innovation that is there only to obfuscate and allow what is essentially theft? Some people want to put a tax on all financial transactions. But that seems too blunt a tool for correcting the actual problems.

Second, here's a tax proposal:
It also means that we can’t go back to tax policies that didn’t generate faster economic growth or jobs, but rather increased inequality. Instead of going backwards, we should adhere to principles like the Buffett Rule, which states that those making more than $1 million should not pay a lower share of their income in taxes than middle class families. We should also end unnecessary tax cuts for the wealthy, and return the estate tax to what it was in 2009.
We know what he means by "tax policies that didn't generate faster economic growth." He told us that those were the Bush tax cuts. It's not clear whether he wants to let the Bush tax cuts expire, or just have the top marginal rate revert to its pre-Bush era value. The difference matters, particularly for how this is sold politically.

It seems clear that Krueger knows the economics literature, as he should. The interpretation of some of the evidence is stretched, though.

32 comments:

  1. I think you mean "...if someone took bananas away from B and gave them to A."

    ReplyDelete
  2. The immigration of tens of millions of low skill imigrants with less than an 8th grade education with the next generation typically dropping out without a high school education -- this plays no role?

    ReplyDelete
    Replies
    1. It may not make much difference whether the low-skilled people are in the country or not for the wages of low-skilled people in the US. Capital is mobile. You can produce the goods here with the unskilled labor, or you can produce the goods abroad and ship them here. It would certainly reduce the wage gap if we let a lot more high-skilled people in.

      Delete
    2. Facts matter. What jobs are being done, in what sectors?

      Can the services and goods provided by those doing construction and day care and landscaping and gardening and housekeeping and janitorial jobs be produced abroad? These are only a handful of examples which could be multiplied.

      Delete
    3. "It would certainly reduce the wage gap if we let a lot more high-skilled people in."

      Let me translate this for young people. Go into debt for $300K to get an education at Wash U, from an enemy like me.

      Then, get your tires slashed when we let an engineer into the US from India, such that your wages are cut.

      IOW, you are really screwed.

      And people wonder why: (1) American engineers and scientists send their sons and daughters to law school; and (2) why most American student with any brains no longer studies science, math, or engineering.

      Delete
    4. Actually you wouldn't go to Wash U to take engineering, unless perhaps you are interested in biomedical eng. My son takes engineering at Carnegie Mellon.

      Delete
    5. Addendum,

      In general, it looks like you have a pretty dismal view of immigration. I'm sensitive to this, as I'm an immigrant myself. Where I come from - Canada - they have a pretty sensible immigration policy that selects for people with high human capital. Canada lets in more people every year per capita than the US does, and it does not appear to have hurt Canadians. In fact, you can make a good case that the benefits have been large. Of course, you could say that I'm here and not there, and why don't I go home, etc., etc. But, the fact is that most people in Canada choose to stay there. No one is suggesting that we have to build a fence on the 49th parallel to keep out the Canadians.

      Delete
    6. "Anonymous" on your comments section is many different individual writers ...

      Delete
    7. In this case, the views seemed to be the same.

      Delete
  3. It is well-known as a theoretical proposition that the income tax has positive insurance effects. The key question, though, is how large those effects are. I think one can make a case that the insurance effects are large, particularly in the long run. When Williamson points to the paper by Manuelli et al. we know that's not serious evidence, as he cherry picks the evidence and papers ignoring the literature that finds large permanent income risk, Davis, Jacobson et al, Krebs etc. etc. etc. which all imply substantial scheme for redistribution which questions the term " going off the rail " and "we know that's no serious
    evidence" it seems that williamson knows the economic literature as he should but the interpretation of some of the evidence seems a bit stretched

    ReplyDelete
    Replies
    1. Well, you seem a little put off.

      1. You could certainly argue that the progressive income tax gives some insurance. Whether that's the best way to provide social insurance is another question. Of course we have other schemes - unemployment insurance, AFDC, social security - that play a related role, and maybe do it better.

      2. Manuelli/Seshadri/Shin isn't cherry picking. Two of those guys are my colleagues, and the other I know well. I can actually vouch for the quality of their work. Trust me.

      Delete
  4. Stephe
    I trust you knowing some myself. The point is not to discard their views as non serious. but I know some of the above quoted guys also and think they are alsoserious. So is Saez i guess though you might not like his non-existing identification strategy and the funny static nature of his model. the point is that there seems no clear cut evidence at all on either side, so based on what economic people should a politician make his decision? why not on saez, diamond, why williamson or manuelli or x, with x having 5 A publications to make it less controversial? I dont understand your aggressiveness to some opinions given that there seems no academic agreement on crucial numbers. isnt that a pretence of knowledge (you might not like hayek) but the professions state does not offer currently a clear cut view, so how do you propose we decide? funny robust control?

    ReplyDelete
    Replies
    1. Who's being aggressive? I'm just asking questions, and giving you some other ways of looking at things. I also think things are less up-in-the-air than you're saying.

      Delete
  5. ok, i appreciate very much your questions and looking at things, I guess we simply have a huge disagreement on the kind of knowledge the profession generates, you seem to think that the elasticity estimates are much more narrow, essentially arguing hat saez and diamond do not reflect state of the art knowledge and are outside confidence bands while i might agree on the art part but view their findings well within current economic knowledge, interesting, i guess i need a more precise definition of state of the art and current knowledge

    ReplyDelete
    Replies
    1. Diamond and Saez is mainly based on Saez's dissertation, about 12 years ago, and it is based on a static model. If you are looking for the state of the art you should read more recent papers. Here are two examples (among many)
      http://www.economics.harvard.edu/faculty/farhi/files/insurance_taxation.pdf
      http://scholar.princeton.edu/golosov/files/odt30.pdf

      Delete
  6. i'm sorry, i dont understand how the banana tree analogy is supposed to function as an argument against Reich/Krueger. it seems like more detail is required

    is a two person world really an economy? where does B get the extra time to take care of 15 more trees and plant new ones? how many banana trees does it take to achieve a decent banana living wage? what happens to the bananas that B cant eat? bananas go bad, you know.

    ReplyDelete
    Replies
    1. The example is supposed to make you think about savings and investment.

      Delete
    2. i got that.
      im just saying the banana analogy is obscure and doesnt help make your point: that the given situation of banana consumption and banana "saving" is better than a redistribution would be.

      Delete
    3. Yes, you don't know whether the new allocation is better than the old one, but Krueger seems to think it is. In the example, there are rich and poor. The rich save more than the poor do. And the way I set it up, it's clear how savings gets channeled into investment. Now, what do you think will happen when we do the redistribution? Consider the case where we're going to redistribute bananas now and every period in the future - it's a permanent change in policy. Think about the whole future dynamic path for this economy, and how it's affected.

      Delete
    4. I think liberals are approaching this issue the wrong (liberal) way. I like Greg Mankiw's approach of treating this as a public good issue. Many people enjoy knowing that everyone is afforded by society some minimum level of opportunities and income, relying on charitable contributions is subject to the free-rider problem, so reliance on mandatory taxation according to ability to pay is necessary. The focus should be on defining what the "minimum" should be, which types of public investment (i.e., R&D) are worthwhile, and what we mean by "ability-to-pay".

      Delete
  7. I don't know why is is so hard to say. It was easily said when we adopted the progressive Estate Tax.

    A high progressive tax is about power, who in society has power and who doesn't.

    The entire tone and direction of society is set by the power of government, to whom it is applied, how, and for what reasons. The more a democratic government taxes, the more democratic it becomes, particularly when then is no Constitutional limit on private spending for lobbying or elective office.

    Being a democracy, a really compelling argument based on sound political science principles that we should set taxes high enough so that no one could self-fiance a campaign for dog catcher or hire a k street lobbyist.

    To illustrate, How would our society be different, if elected officials were paid based on "real" increases in income and wealth of the lowest 20% of our society? We all know that the policies of the gov't would be radically different.

    ReplyDelete
  8. "I think one can make a case that the incentive effects are large, particularly in the long run."

    I didn't find your case for large incentive effects particularly persuasive, but I got into that on that thread. But a comment about the "in the long run" part... that's a good reason why we should focus on taxes on the rich as the way out of our current budget situation, with the economy as fragile as it is, rather than taxing the middle class or cutting spending, which would have much more immediate negative consequences.

    As for the long run, it seems to me that there is a long run problem with allowing inequality to continue to grow over time. The wealthiest tend to consume less and seek out interest earning savings more. To some extent this can be good because it becomes easier to get credit and more investments are made. However, as inequality continues to grow the increasing savings at the top chases lower and lower quality investment opportunities, the financial sector goes to greater lengths to lure people into taking on debt, the stock market gets bubblier because price-to-earnings ratios rise as there is more money looking to purchase stock but not proportionally more money for consumption to create earnings... well, that second part can be avoided for a while thanks to increasing debt, but eventually, if wealth continues to be increasingly concentrated in a small number of hands, you're going to end up where we are today.

    ReplyDelete
    Replies
    1. In economics we learn that there is never a free lunch. Every good thing comes at a cost. In this case, you can't redistribute without incentive effects. If you try to give the people with smaller pieces of pie larger pieces, the total size of the pie will shrink. The key questions are: (i) What's the quantitative nature of the tradeoff? (ii) how do you do a given redistribution to minimize the incentive effects? (iv) What is actually politically feasible?

      Delete
    2. What's the opposite of a free lunch? An expensive bag of ****? Situations where people lose and hardly anything is gained? Are there any of those in economics? Because unless those are hard to come by and we just don't have any of them, there must necessarily be free lunches as well, in the form of disposing of expensive bags of ****. When we lost big in the economic crash, was did someone else equally gain? I believe you are a monetarist, so when the Fed reacts to a recession by targeting lower interest rates, who are they screwing over, and why do you support screwing them over and for whose gain? Or is that a free or at least cheap lunch?

      Incentive effects I'd expect are mainly a relative issue; for example, raising the income tax on the rich would encourage them to prefer to invest their money in growth stocks rather than dividend stocks, but that distortion could be corrected by raising the capital gains tax with it. Unless you go with a very steep tax rate pretty much no one is going to react to having less money by working less; income effects are the primary effect of taxes, but if you're far from living paycheck-to-paycheck that is a long term effect.

      And then there are the real (non-money related) constraints on goods and services the economy can produce -- I assume that's what you mean by economic pie. But why must the pie shrink? The pie isn't something that exists in a certain quantity, it is made, and we have an unusually large number of people not making any of that pie right now, so it stands to reason there is less pie to go around than there could be. Who must quit their job if we give a teacher their job back?

      Let's say I have an economy with three people in it, Alice, Bob and Carl. Alice is a wealthy financier, while Bob and Carl are both unemployed. (If you need to for plausibility, add a few more people to the economy to produce the goods Alice wishes to consume and to produce for each other.) Bob and Carl have unmet needs, and they each have the skills to meet each others' needs, but they have no money to pay each other, so they remain unemployed. They also owe money to Alice. Alice has no trouble getting whatever she needs without employing Bob and Carl, but she is concerned about her investments because some of her debtors are broke and are missing payments. Now let's say a meddling government steps in with a redistributive tax causing some of Alice's money to be given to Bob and Carl. They then each give money to each other in exchange for some service, then give the income they earned from their work to Alice. Alice is no longer stressed about the payments that aren't coming in; she's grumbling about taxes instead. Bob and Carl had real gains in their welfare, that resulted from them working to create valuable goods or services that otherwise would not have been created. What was lost? Does anything like this situation ever happen in the real world?

      Delete
    3. Pretty poor effort -- you're saying that Bob and Carl, despite each being able to supply a good the other wants and knowing that, cannot agree to exchange those goods directly. Lies make Baby Jesus cry.

      Delete
    4. Okay, make it Bob, Carl, and David, and Bob needs something Carl can provide, Carl needs something David can provide, and David needs something Bob can provide. Or, they may not know each other; they may only know the businesses they used to work for that might hire them back if they needed more workers to keep up demand. There's a reason we don't do much business by barter in the real world; the actual trades we want to make are quite complicated and difficult to arrange by barter. I made the economy simple for clarity, but my point doesn't depend on an the economy being so simple that barter is a practical alternative to money.

      Also, consider learning the definition of the word "lies" before using it again.

      Delete
  9. What about income inequality and crime?

    Any research in that area?

    ReplyDelete
    Replies
    1. I don't know, but it should matter. Education is much cheaper than incarceration.

      Delete
  10. I do think that there are negative aspects of inequality that are not taken into account of the studies you mention.

    Indeed, it is an incentive device but if inequality has a positive impact on crime, the costs of "a too high level" inequality might outweigh the benefits. As you said "Education is much cheaper than incarceration." Of course, what is then the "optimal level" of inequality.

    Also social mobility between income groups might be dampened by an excessive level of income inequality, in particular for the poorest (probability to moving up very low) and the richest (probability to moving down very low). I have in mind the NYT article that showed the probability of remaining in the lowest/highest income group is much higher in the US than European countries. My question would be: does income inequality prevent social mobility? Again, I do not know any studies relation these variables although I do think it is relevant in this context.

    ReplyDelete
    Replies
    1. I think the important question relates to what you propose to do about the income inequality. People seem to be focusing on redistribution through progressive taxation, but why not pay more attention to education and health?

      Delete
  11. To see how much education affects inequality, I think it would be useful to characterize a "a natural level of income inequality". Assume that this level is equal to "ability - or - productivity" differences across agents. Suppose that it is possible to identify the distribution of ability. Then, one can study how a better education policy, i.e. one where everyone has ex ante the same opportunities ( i.e. independent of their parents wealth), may effect inequality. In that sense, I think the crucial part would be to identify the "natural level of inequality" and then one can think about education and redistribution.

    One may then be able to and the resulting income inequality

    There is an issue with education. Assume that inequality should reflect "ability differences" across agents, then even if we have

    If ability across agents

    ReplyDelete