Tyler Cowen on the Past and Future

Tyler Cowen gave a talk at UNC today, in what appeared to be a class taught by Mike Munger. It was open to the public, so I went. Cowen introduced it as a mix of his last book and what will become his next one. The former considered micro and micro-micro economic development; the latter concerns the macro economy. He closed by discussing implications for the American political economy.

Cowen began by arguing that the most notable economic developments in the U.S. in recent times has been the ability to collect and manipulate information, especially what he calls "cultural information". The ability of individuals to collect and readily access culture at very low marginal cost through social networking and digitized media has allowed us to create our own economies. These have dramatically improved our quality of life, but unlike previous inventions they have not increased GDP by much or employed many people.

Then he shifted to his macro view, which is most heavily influenced by two events: the stagnation of median wages since 1973, and the financial crisis. The former indicates, to Cowen, that we haven't been as innovative as we thought. Most of the important inventions (which he loosely defines as mixing fossil fuels with machines) occurred well before 1973, and we've spent the time since making marginal improvements to the same technologies. As we've done so we've increased productivity, and that's why everyone has a refrigerator and a telephone. But we haven't really come up with new innovations; we've just improved the old ones. The exception to the rule -- information technology -- has improved quality of life but not measured GDP.

Cowen brings this together by saying that he is a "utility optimist" but a "numbers pessimist". He thinks that we'll continue to improve the ways we can collect and manipulate information and this will have important real benefits for people, but they will not create many jobs or provide a large boost to GDP. He says that we cannot expect to maintain a trend rate of real GDP per capita growth of 3% a year; 1% -- which is more than what the median earner has had since 1973 -- is the new normal. That doesn't mean we're stagnating; it just means that we have poor measures of progress. He recommends the Wolfers/Stevenson happiness research as an ongoing attempt at correction.

But this divergence between numbers and utility is where he sees the problem for political economy. Voters will demand 3%, rather than accepting 1% plus non-monetary improvements in standards of living. Politicians will thus promise 3%, and will pursue policies that generate it. That means encouraging a debt-based economy, encouraging too much consumption, and encouraging bubbles in asset prices that lead to financial crises. He didn't explicitly say it, but it sounds like he expects boom-and-bust cycles to continue until the American public is willing to accept 1% growth, or until we break through the "innovation plateau" that we've been stuck in since 1973.

I think I've summarized his argument correctly, and I'm sure he'll be writing much more about it in the future. I think it's a compelling story, but I'm not yet completely convinced. Here's how I see the world since 1973:

1. The natural advantages of the U.S. economy post-WWII had mostly dissipated by 1973. This was inevitable, indeed it was something the U.S. strove for, so the previously-high growth rates were simply not sustainable. This isn't about innovation; it's about competition. As W. Europe and and Japan "re-industrialized" and were able to productively mobilize labor, they narrowed the U.S.'s margins. At the same time, the U.S. had mostly already reaped most of the GDP benefits of mobilizing female workers and integrating minorities by 1973.

2. Somewhat related to #1, I think Cowen has the wrong level of analysis. While it may be true that median incomes have stagnated in the U.S. since 1973, real global GDP/capita has nearly doubled since 1973. Even allowing an increase in inequality, global median incomes has certainly increased markedly, probably well more than 3% a year. (A quick search didn't turn up a global median income growth time series, but I can't imagine this isn't true.) We would expect this to happen as more countries employ their populaces in industrialized work. In other words, the experience of the U.S. from 1900-1973 has become the experience of the world from 1973-2010. This has put pressure on American middle class wages, as we should expect it would: when the supply curve of less-skilled labor shifts right, the returns to less-skilled labor goes down. But the returns to more-skilled labor have not gone done, which is why the American mean and median have diverged.

3. I don't think the new normal has to be 1% growth. It could also be 3% growth, but not broadly dispersed. That has, in fact, been the story of American post-1973. Not all of that growth was a myth. After all, before we got the micro-micro innovations like Twitter and iTunes we also got the micro-macro innovations like the PC. These did raise the real productivity of the economy, but not necessarily for the factory worker or custodian. Those initial innovations made Bill Gates much richer in monetary terms than Joe the Plumber, but Joe the Plumber got psychic benefits that he would not have otherwise had. In Cowen's language, numbers and utility went up, but not necessarily in equal amounts for everyone. I don't see that that process has run its course. Facebook and Twitter may hire many fewer people than GM and Ford hired when they boomed, but Mark Zuckerberg is the youngest billionaire in history.

4. If #3 is correct, then the political economy dimension becomes about distribution of monetary gains, not divergent perceptions of utility vs. numbers. Interestingly, this will not be a battle between capital and labor, but between labor and labor. (It doesn't take much capital to create Facebook; Zuckerberg did it in a few months on an IBM.) Maybe labor becomes more of a lottery. If that continues to happen, I'd expect the political equilibrium to be a strengthened welfare state. I don't think recent political swings necessarily contradict this, since the best political models are the most structural political models. Additionally, public anger over the bailouts is much stronger and deeper than many expected, and no one is interested in weakening the major entitlement programs.

I'll have to think about this more for than just an afternoon/evening to form stronger conclusions, but that's where I'm at now. Perhaps as Cowen develops his thesis more fully he'll address some of this, and especially consider if/how the story changes when we think globally. Either way I'm looking forward to seeing how his thoughts develop.

P.S. I live-tweeted the lecture, and asked the Twitterverse for questions. Daniel Davies asked me to ask Cowen what he thought of Keynes' "Economic Possibilities for Our Grandchildren". Cowen's response was essentially "It is one of the more interesting intellectual mistakes of the 20th century." I don't think Davies liked that response too much, but I'll let them speak for themselves if they like. I'd never read the essay before. It is interesting. And it is, I think, mistaken, although not entirely. A pdf is here.

I Don't Think This Means What You Think It Means

Trade skeptics Eyes on Trade noticed that wages decline during a recession:

On Nov. 3 the U.N. agency on labor, the International Labor Organization (ILO), released a 15-page report finding that real wages fell in countries around the world, including the U.S. and some other wealthy nations, raising questions about whether workers are sharing in any global economic recovery.

The report included data from 35 countries, and found that monthly wages have fallen almost 2 percent in the U.S. since January 2009. The ILO found that inflation-adjusted wage growth fell sharply around the world in 2008 to 1.4 percent, down from 4.3 percent in 2007, and wages continued to fall in a number of countries in 2009.

This continuing drop in real wages around the world illustrates the need for trade policies and agreements that protect workers’ rights and prevent a further “race to the bottom” in global wages.

So first of all, this means that real global wages actually increased in 2008, though at a slower rate in 2007. And it is not surprising that wages fell during a global recession in 2009. Why? Well, what is a recession? It is a drop in economic output. In other words, the world produced fewer goods and services in the first half in 2009 as it had before. The IMF projects that global GDP for all of 2009 will be negative 1.1%. When output goes down, employment goes down. When employment goes down, so do wages.

More importantly, this has nothing to do with a "race to the bottom" in wages. Wages aren't down because states are trying to lure foreign capital by weakening standards; wages are down because economic activity is down, full stop. A "race to the bottom" story would have to show that real wages declined even as output increased. The U.N. report cited by EoT contradicts this story by demonstrating positive wage growth in 2007 and 2008 (years in which output increased), as does this ILO report (pdf).

The truth, of course, is that "race to the bottom" narratives are too simplistic: exposure to trade affects different states differently, affects some standards differently than others. In most cases, trade exposure improves labor standards (perhaps through a "California effect"). But in terms of wages, on average workers in export-oriented industries get higher wages whether they work for locally-owned firms or MNCs. (Dr. Oatley discusses this on pp. 366-372 of his textbook.)

Stagnant Incomes for the Middle Class?

Common wisdom tells us that incomes for "the rest" have failed to rise much in the last thirty years, while incomes for the lowest quintile have actually lost ground. Recent research by Terry J. Fitzgerald, senior economist at the Minneapolis Fed, questions this common wisdom and reaches some quite different conclusions.

In a nutshell, Fitzgerald finds that "labor income per hour for middle America has not stagnated. Rather, the economic compensation for work for middle Americans has risen significantly over the past 30 years."

It seems that the conclusion one reaches about real income growth rest heavily upon what price deflator one employs and how one treats non-wage benefits. If one uses a consistent deflator and includes non-wage fringe benefits as part of compensation, real hourly wages have risen by 25 to 30 percent since 1975. Not huge gains over thirty years, but also not consistent with the commonly-state claim that incomes are falling.