Krugman deceives Yglesias (updated)

The gullible Matthew Yglesias defends Krugman's attempt to cheat:

"In general, rich countries normally stay rich and poor countries normally stay poor."

Theory and a mass of empirical evidence show that in general, richer countries grow slower and poorer countries grow faster (this is, by the way, simultaneously true with richer countries remaining richer. I will let clever Yglesias figure out the false dichotomy himself) .

The fact that America has not only kept its advantage over Europe but also even expanded the gap is extraordinary. The US is the exception, the only country in the top five 100 years ago that is still in the GDP top five (and yes, the snide remark aside, the US in 1900 had far more pro-market institutions than Germany or France). And even the gap between the US and Western Europe has closed in the last 100 years, it only stopped converging in the 1980s.

The pattern during the last 100 years has been for countries with functioning economies to grow faster the lower they start. If Yglesias was an trained economist he would know all this. Conditional Convergence is one of the most robust relations in growth theory. Krugman knows this, but Krugman is a liar, he just wants to maximize his ideological argument at any given point, deceiving his readers if he has to.

That's why Krugman has not explicitly made the claim that growth rate and levels are unrelated, he just ignored the levels altogether. He knows that writing what you wrote would make him look foolish to other economists. While Krugman knows what he is doing (he just doesn't care), my guess is that Yglesias is honest in his argument, following Krugman where Krugman himself carefully does not step. Matthew Yglesias is left standing with his naive post on growth theory, his just reward for trusting Paul Krugman.

Allow me to demonstrate. Krugman compares per capita growth figures for Europe and the U.S. between 1980 and 2008. I have plotted the per capita growth rate 1980-2008 and starting per capita income in 1980 for available OECD countries (Eastern European countries and unreliable Luxemburg are excluded). The U.S and E.U.15 are included in the plot, but NOT included in the simple regression line (the individual members of the E.U.15 are already there).


See any pattern?

Note that the US performance is above the linear prediction and the E.U15 slightly below. This is not in any way scientific, but using the simple linear framework the US, starting above average, "should" have grown by 46%, and the EU.15, starting around average, "should" have growth by 74%. In fact the US grew by 71% and the EU by 66%.

Here is another one with even more countries (more data available, I threw in Luxembourg so no one will accuse me of cheating, although their GDP figures are not reliable) from 1995-2008.


Still care to deny the relationship between growth and level?

Here is the same logic, applied to U.S states. There is only data available from 1990, so here I have plotted 1990-2008 growth and starting level in 1990.


Again, notice any pattern?

American states that started of richer in 1990 have grown faster since.


I pointed out yesterday that the E.U 15 has the same per capita GDP as Alabama. Being a poor American state, Alabama grow faster than the US average between 1990-2008. In fact it grew by 1.75% per year. During the same period the EU.15 managed to grew at 1.64%. (in other words Alabama per capita earnings grew by 37%, the E.U 15 by 34%). In the same period rich Maryland grew by only 1.39%. By Krugmans "logic", Maryland should be learning from Alabama.

Alabama has the same per capita income and slightly faster growth rate as the Social Democratic EU.15, which Krugman wants us to believe is a "Dynamic" region that the US should "learn from". Has Paul Krugman ever written a column asking us to learn from the economy of Alabama? Of course not. That would be simply idiotic. Alabama is poor, and has a lower standard of living, just like the E.U 15. It only manages to grow faster than others because it starts off at such a low level (the EU doesn't even manage to do that).

Being poor and growing at average or slightly above average is not something that anyone (at least anyone not driven by dogmatic liberal ideology) would recommend as a shining example for others to emulate.

What is most scary is that the left either denies the American advantage or seems to think that the wealth advantage of the US is somehow god given. It is just a historical coincidence that the capitalist pro-market economy does better than the more leftwing nations. They can make American policy Social Democratic without making US economic performance Social Democratic.

PS. As a dysfunctional part of a well functioning economy, Alabama has more social problems with the same level of income compared to Europe. The quality adjusted standard of living is certainly higher in Europe, for a given level of GDP. On the other hand the standard of living of low income rural American states with small social problems such as Maine is probably higher than Europe, for a given level of GDP.


Additional data:

Some people in Marginal Revolution are denying conditional convergence in income. In order to convince more readers,here are the p statics, R-square and correlation coefficient of simple linear regressions of starting per capita GDP and average growth, so you are not forced to rely on your eyes:

Growth 1980-2008,

24 OECD countries:

p=0.003

R-square=0.331

Correlation-coefficient= -0.575

Growth 1995-2008,

35 OECD countries:

p=0.000

R-Square=0.322

Correlation-coefficient= -0.570

Growth 1990-2008,

50 US states:

p=0.000

R-Square= 0.4206

Correlation-coefficient= -0.6485

I think most economist would consider a correlation of -0.57 pretty decent.

Remember that Conditional Convergence is not something I made up, it is the dominant theory with strong empirical support among the OECD countries (and regions within rich nations).

Krugman cannot be allowed to make up economics as he goes along, asserting whatever suits him that day (if he does, he better have data on his side).

Here is Bob Lucas, in the Journal of Economic Perspectives:

"the model fits the fact that in the postwar period growth rates vary much less among the advanced economies than among the poor and middle income economies. It presupposes the existence of an ever-growing "convergence club": a set of rich economies within which income inequality is falling, even in a world in which overall inequality is rising or not changing very much. It can be interpreted as implying a focus on conditional convergence, since conditional on both having left the stagnation state, any two economies are getting closer to each other."

Robert E. Lucas (2000). "Some Macroeconomics for the 21st Century" Journal of Economic Perspectives vol 14(1)

Here is a picture from a recent Lucas paper, for open countries:

Robert E. Lucas, (2009). "Trade and the Diffusion of the Industrial Revolution," American Economic Journal: Macroeconomics, 1(1)

Additional data and argumentsII:

I replied to Yglesias in the Marginal Revolution comment section. He wrote:

"I can't even tell what this argument is about anymore.

Initially, Jim Manzi said that Europe had grown more slowly than the US in recent decades, and that this shows that social democratic policies are bad. Krugman observed that it's not true that the US has grown faster than Europe.

Then Greg Mankiw came along and observed that some European countries with similar growth rates to the US, nonetheless have lower per capita GDP levels, implying that this is the economic problem with social democracy.

So I observed that the per capita GDP gap between the US and countries like Germany and Italy long predates the emergence of post-WWII social democracy.

Now you're "rebutting" this, I guess, with references to the conditional convergence literature. But what does this rebut. Go back and read my post and you'll see that, exactly as you're saying, the US-Europe gap was bigger in 1900 than it is in 2010. This makes it hard for me to see evidence for the claim that the adoption of social democratic policies in the postwar era can be to blame for European countries having lower per capita GDP than we have in the United States. The existence of the gap is longstanding and has declined over time even as a substantial "leisure gap" has opened between the US and Europe.

I note that this whole debate could be made much simpler by people saying "what Jim Manzi wrote was wrong, but I still think there are other reasons to believe that free market economic policies are superior.""

I wrote:

"Mattew,

1. You are making erroneous claims to begin with. Krugman did not only question Mazis claims about growth. He went much further. Here are Krugmans claims:

" the image of Europe the economic failure is so ingrained on the right that it’s never questioned,"

" the story you hear all the time — of a stagnant economy in which high taxes and generous social benefits have undermined incentives, stalling growth and innovation — bears little resemblance to the surprisingly positive facts. The real lesson from Europe is actually the opposite of what conservatives claim: Europe is an economic success, and that success shows that social democracy works."

If you want to evaluate "economic success", you simply cannot rely entirely on growth rates. Levels are what we ultimately consume. By that measure Alabama is an economic success story and Delaware is economic failure.

Krugman pay no heed to the fact that Americans each year, each month, each day produce 36% more than Europeans.

The difference between levels is 8 times as high as all the yearly output lost in the recent crisis. This makes Europe an economic "failure" when compared to America.

And if Krugman thinks a 36% is too low to count as success, he should explicitly say that. Instead he simply IGNORED the most relevant figure for his claim.

2. Krugman's argument ignores a salient fact about growth theory, that with functioning economies poorer countries grow faster that richer ones. This has been the trend the entire century. When it comes to America and Europe, this has been true for the entire post-war period, up until the 1980s.

Here something strange happened. Not only did Europe stop naturally (according to growth theory and historical patterns) converging, it started to lose ground. the US grew 6% faster than Europe, when the lower starting level of Europe suggested it should grow.

This phenomenon has been noted by economist, and policy makers in Europe. They are discussing ways to turn it around (the Lisabon Strategy), with explicit reference to this pattern.

Instead Krugman pretends the entire controversy is due to faster population change. But of course economists and policy makers across the world are not stupid, they knew how to adjust for population growth. The striking fact was Europe stopping to converge in a per capita basis.

Note, the pattern of convergence did not stop. Within Europe and for other OECD countries it continued. Only America was doing better than it should have based on levels alone, and EU.15 worse.

3. Up until now I am pointing out distortions so blatant that even the left cannot deny them. Now a subtle argument, that I don't really expect non-economist liberals to accept:

The central debate is about the effects of Social Democratic policy on living standard. Economic theory basically predict that your level increases each year with technological improvement, and also converges in response to exogenous changed.

Once the convergence is finished, the country will again grow at the rate of technological improvement, but with a different level.

If we for example believe that high taxes reduce hours work by 10% and that this is reflected fully in GDP, the effect of a tax increase is to first lead to a few years of reduced growth (lower than what it should be, not necessarily negative) and, once the economy is 10% below where it would have been, to AGAIN start growing at the rate of technological change.

Do you see now why Krugmans comparison of growth rates is deeply misleading? Once the effects of Social Democratic policies have taken hold already (they certainly have had plenty of time), there is no reason we should expect Europe to grow slower. But that does not mean they had no cost! They still produce less, year after year.

The standard theory of the right is that taxes reduce living standard, reflected in a few years of lower growth followed by a lower level. If bad American policy also reduced the long run rate of technological increase it will do so by the same numbers for the US and Europe alike.

This debate would be much simpler if Krugman acknowledged that the policies he proposes would, as a best guess compared to Europe, lead to a 35% decline in American per capita income. Once his policies have taken their toll on the economy the growth rate would return to normal, but at levels forever below where they would have been."

Funny that Yglesias does not link the fact that the "leisure gap" opened precisely at the point in time when Europe started to tax labor, subsidize not working, and stooped growing.

Let me just add for the sake of being pedantic that unlike what Yglesias claims the US-Europe is not smaller now than in 1900. Partially because of Europes weak performance since 1980, the gap between the 12 western European countries is actually slightly bigger now than 1900. I don't want to push this mistake, since it is not central to Yglesias argument (he looks at Nationmaster, he should have looked at the original data by Maddison).

A few more world on convergence, while we are at it. Here is western European per capita GDP as a share of US, from 1950-2006.

Notice the early consistent convergence. At some point in the 1970s or early 1980s the convergence stops. I use 1980 here, the year Krugman chose to prove Europe was a success story.

There are two leading possibilities. One is that the US is naturally better at making goods and services than Europe, and that the earlier convergence was merely recovery from WWII.

The other possibility is that it reflect policy differences, with the convergence stopping a few years after Europe started to diverge from the US. Social Democratic Europe is "stock" at permanently below the US (or until the US also becomes a Social Democracy).

I just want to show you another picture, to remind people about how "normal" conditional convergence looks like. Instead of comparing western Europe and US, it compares outlier Switzerland with western Europe in the post war period. This is what Europe's experience compared to the US should have looked like, if we believe Americans are not inherently superior to Europeans (I certainly don't), and if there were no policy differences between the two regions.

Looks different, doesn't it? The question Yglesias should as himself is why it looks different. Just coincidences that we can ignore for the sake of discussing policy, or are deeper economic forces in play?