« YouTube and Spontaneity | Main | Why Immigrants Run Bodegas »

August 21, 2006

Who's To Blame For Inequality?

There's a fun argument rushing through some blogs today sparked by a Paul Krugman column on the government's impact on inequality.  Krugman argues that the correlation between conservative and progressive political periods and rising or lowering inequality is too powerful to ignore.  "it seems likely," he writes, "that government policies have played a big role in America’s growing economic polarization — not just easily measured policies like tax rates for the rich and the level of the minimum wage, but things like the shift in Labor Department policy from protection of worker rights to tacit support for union-busting.  Brad DeLong disagrees, saying that "the shifts in income inequality seem to me to be too big to be associated with anything the government does or did."  And Matt Yglesias disagrees with Brad, saying "The trend data is too striking to be ignored. If you have a phenomenon and are having trouble identifying the cause, the thing to do is to try harder to identify the cause, not assert that the phenomenon isn't happening."

Let's start here with what the trends in inequality actually look like.  Here's a graph tracking the gini coefficient -- the standard measure of inequality -- over the 20th century.  The higher the coefficient, the more inequality:


Okay.  Now, the problem Paul and Matt are having is that nothing the government actually does looks able to generate such wild swings in wealth distribution.  That's why Brad won't buy their argument, though he doesn't provide an accounting of what does account for the shifts.

Seems to me we have a causal problem here.  Politics, after all, tends to follow societal trends, not the other way around.  So during periods when conservatism is ascendant across the country, conservative politicians will win elections.  But it's probably the grassroots sentiment, not the legislation, that accounts for much of the difference in income distribution.  During these periods, support for unions (and thus their strikes), will weaken, pressure on corporations to ensure wages keep pace with productivity will essentially evaporate, CEO pay will skyrocket, etc and so on.  These trends aren't created by the regime in Washington, but they're enabled by conservative politicians who won't attempt a governmental correction, and preserved by a populace that, for whatever reason, is unwilling to take a populist turn. 

Eventually, though, that changes, and corporations have to react, and unions win some strikes, and the general trend is towards higher salaries and better benefits and whatever else, and during all this some progressives get elected to office, and so we assume it's all their fault, when all they did was raise the minimum wage fifteen cents.  But because we can see which party controlled Congress easier than we can track societal attitudes, we tend to blame the shifts on political changes without knowing exactly why.

That, of course, is not to say that there's nothing the government does.  Whether tax rates encourage inequality, they can certainly discourage it, and, for a very long time, did so.  Here's a graph tracking the top marginal tax rate through the 20th century:


Until the Reagan era, the very richest would see the top fractions of their incomes taxed at rates of 70 (and, at times, 90) percent.  That exerted a fierce check on how rich the very richest could get, curbing inequality even if it didn't necessarily make the middle class or the poor much wealthier.  The Bush tax cuts, galling though they were, had nowhere near the effect of Reagan's rate slashing.  So too are unions, corporate regulations, wage standards and subsidies, and the generosity of redistributive programs important for inequality.  All those are government actions though I'd argue they're all secondary to decisions made in the private sector, and done in reaction to public opinion. 

It's also worth saying that government is far more effective as a check on inequality than as an accelerant.  Various trends, some pernicious (corporate greed, union decline), some not (technology, globalization, single mother families), contribute to inequality.  What government can do is tax and redistribute in such a way that growth is shared equally across society.  During conservative moments, it doesn't even make an effort to do that, and society is the worse for it.

August 21, 2006 | Permalink


So, it seems that another depression is needed to get things back on track...

Posted by: George | Aug 21, 2006 11:21:27 AM

why are you discussing this- and not Jon Benet?

Posted by: akaison | Aug 21, 2006 11:30:38 AM

Please add some detail to your graphs. Viz: the date (33?)of the Wagner Act which allowed unions to form and to win strikes; The date (53?) of the Taft-Hartley act which took away the ability to organize; The rule changes in the seventies which removed the right to strike; and the destruction of Patco by Reagon in 1980. Then add a graph which shows the per centage of the population with Union Membership. Government actions encouraged and then discouraged Unions in step with your graphs.
For more see "Which side are you on?" by Thomas Geoghegen http://www.powells.com/biblio/7-1565848861-2
-- ml

Posted by: Martin Langeland | Aug 21, 2006 12:39:10 PM

It's not just taxation, it's also union policies (see Kevin Drum on the same subject). I suspect Brad is right that trends other than governmental action started the increase in inequality (e.g. globalization and the winner-take-all phenomenon), but the actions of the US government since 1980 have done nothing but accelerate the trend (union-busting, deregulation of corporations, the shrinking minimum wage, and of course tax cuts). This has been done at a time when government, in the interests not only of fairness but also social stability, should have been working in the opposite direction. What's happened is that as the rich have become the super-rich, their ability to capture the political system and bend it to their ends has also geometrically increased.

Posted by: Rebecca Allen, PhD, ARNP | Aug 21, 2006 1:56:51 PM

Brad DeLong's framework is a simple one lifted from an elementary economics textbook: in that model, a firm's output is a mathematical function of factor inputs; income derives from selling output at market price, and the income of the factors divides according to the marginal product of each factor.

This theory is wrong in more ways that can be conveniently counted in a blog comment.

A more sophisticated analysis of income takes into account that market outcomes are continuously variable and uncertain. Ford Motor Co. can only estimate how many cars it will sell next month and the actual number varies randomlys over time; ditto, for supermarket, the shoe shine boy, the doctor, and other actor in the economy. All market outcomes are contingent, and therefore all incomes are contingent.

Firms try to manage their operations to better utilize the capital resources they own, so as to increase the income earned by those capital resources. At the time of product sale, almost all of the associated capital investment is a sunk cost investment, and firms must seek market power to ensure that they are able to recover quasi-rents on their capital investments. In contrast to Brad's ludicrous theory, compensation to labor is paid not for a monolithic "unit" of labor input, but to induce desired behavior. In many cases, the firm is committing to pay, say, a salary, regardless of market outcome (short of having to lay off the worker), and then managing a production process to realize a residual of the owners of the firm.

I don't expect anyone to really understand any of that from a blog comment, but people, who have taken some economics should recognize the gist.

The point is income and risk are inextricably linked. What FDR did in the New Deal was not follow the zeitgeist; what he did was 1.) build out a lot of infrastructure, which expanded the reach of electricity and roads (and therefore the industrial revolution) geographically, and 2.) add a lot risk-dampening, insurance to the economy.

Ordinarily, a deflationary Depression puts a lot of risk into the economy. A lot of people at the bottom lose their jobs, their homes, their businesses and their savings. People at the top end up owning all that stuff. The original gilded age of the late 19th century happened exactly that way. A prolonged deflation was used to drive down wages and wipe out banks (and people's savings) and mortgaged farms and unions. The economy grew rapidly, because it was the time when the West was being developed and railroads and electricity and large-scale industry and oil and steel were all expanding, but almost all the growth benefitted a very few, as waves of bank and business failures wiped out everyone else, over and over again.

FDR reversed the usual outcome of a deflation. He instituted social security, unemployment compensation, deposit insurance and bank regulation, the SEC and securities regulation, a system to promote low-interest mortgage lending, promoted unionization and small business cartels, etc., etc.

All of those institutions and policies have to do with providing cheap insurance to poor and middle class people. It changes how those people make key economic decisions, like whether to save, whether to buy a house, whether to send their kids to college, whether to start a business, whether and how long to search for a better job. And, they affect how likely it is that poor and middle class people will be able to hang on to whatever economic achievements they are able to secure. Like the lady in Fried Green Tomatoes, having more insurance makes you more powerful.

The two political parties divide precisely over the general issue of how much social insurance to provide. So, when one Political Party prevails, it is like a wind blowing over a field of grass. Every government policy is bent and reformed in a one, general direction. And, the predictable outcome of reducing social insurance is to increase inequality of income.

It is not just unionization, it is a matter of how easy it is to get Social Security disability, how vigorous is the SEC, what the rules are on predatory consumer lending or personal bankruptcy, and on and on. A whole field of grass -- the whole web of policies and rules and regulations that limit and channel and offset risk. Republicans want more people exposed to more risk; they argue that it sharpens incentives and expands productivity and the economy generally; it also, predictably, can grind down on people, causing them to be either more risk averse or more desperate in their behavior, to live paycheck to paycheck, to be afraid of losing a job or taking a chance on education, to take on credit card debt, to forego buying a home, etc. etc.

The better analyses of the Social Security reform "personal accounts" proposals highlight this aspect. Personal accounts would be a very good deal for high income, high wealth people, who could invest their personal accounts in high risk, high return investments, "insured" by the existence of an array of other opportunities and investments in their life portfolio. Some poor people might get lucky, but others would lose out big time. A substantial transfer of wealth and income from the poorest S.S. beneficiaries to stockbrokers or other wealthier investors would occur, under almost all of the "personal accounts" plans.

This kind of shading of income and wealth distribution by the imposition of risk figures in most government policies, and the two political parties systematically disagree on how far this should go. So, it should not surprise anyone that when one political party prevails, all government policy and general trends in income and wealth distribution go with it.

You can see the evidence in a lot of economic statistics. Look at the unemployment rate trend under Reagan-Bush, then Clinton, then Bush II. Look at medical cost inflation or pharma prices.

Posted by: Bruce Wilder | Aug 21, 2006 5:25:57 PM

"pressure on corporations to ensure wages keep pace with productivity will essentially evaporate"

That sounds like a very odd method of explaining how pay is determined. Do we actually think that managers raise pay because they "ought" to? Because of social pressure?
I thought it was accepted that companies (at least tried to be) profit maximising institutions. Thus pay is raised when the labour market is tight and it is necessary to attract labour from other companies, rather than being able to hire from the pool of the unemployed (or, if you prefer, to overcome the preferences of some to stay out of the labour market altogether).
A rise on productivity therefore feeds through into higher wages by companies competing with each other for access to the greater profits available from employing that higher productivity labour. This only happens as and when the ecoonomy is growing faster than productivity: which, as many have noted recently, is not the case currently. Productivity is growing about as fast as (real) GDP.
Now that supposition may or may not be correct but it does seem at least more likely thatn the method Ezra seems to be proposing: that wages rise alongside increased productivity because managers are shamed into it by social pressure.

Posted by: Tim Worstall | Aug 24, 2006 6:18:07 AM

This was a really terrific and concise summary of the argument and the positions of the major commentators on both sides of the economic inequality debate. Thanks so much for condensing it down for us and for including the links as well.

Posted by: panasianbiz | Sep 20, 2006 1:19:44 PM

The comments to this entry are closed.