Bryan Caplan  

Murphy's Catch-Up Contradiction

PRINT
Gifts in Kind vs. Gifts in Mon... George Akerlof and Rachel Kran...
Bob Murphy thinks that my recent posts on Soviet and European growth "seem contradictory":
Now I'm not saying Caplan was wrong about the Soviet Union, and I'm not saying this new guy is wrong about Europe. What I am saying is that these two positions seem contradictory. When the Soviet Union grew more slowly than the US, the GMU'ers said, "Ha! We knew it! Only you commie lovers would have used the Solow model on a capitalist and a socialist country." Yet when Europe grew as fast or faster than the US, the GMU'ers said, "So what? Using the Solow model we would expect the capitalist and the social democratic welfare states to show these characteristics."
But things are not as they seem.  If Samuelson and other Soviet growth optimists were just appealing to catch-up, they certainly wouldn't have predicted that the Soviets would surpass the U.S.  My claim wasn't that it was crazy to think that the Soviet Union could temporarily have higher growth than the U.S.  My claim, rather, was that "There was never a point in Soviet history when a sensible economist would have seen communism as good for growth in any meaningful sense."  If Samuelson and company had said, "The Soviet Union has decent growth for the time being because its ability to copy more advanced economies outweighs the harm of its terrible economic policies," my criticism would be less harsh (though I'd still fault them for trusting Soviet statistics, ignoring collectivization famines, etc.).

The same goes for Europe.  Its relatively statist economic policies didn't prevent it from exceeding U.S. growth for decades.  But with better policies, Europe would still be gaining on us - or would have already caught up.


Comments and Sharing





COMMENTS (14 to date)
Boonton writes:

If Samuelson and company had said, "The Soviet Union has decent growth for the time being because its ability to copy more advanced economies outweighs the harm of its terrible economic policies," my criticism would be less harsh (though I'd still fault them for trusting Soviet statistics, ignoring collectivization famines, etc.).

Actually Samuelson did say that in 1976:

"They [referring to 'many experts' from a previous sentence] suspect that, as the Soviet Union reaches stages of development more comparable with ours and places greater emphasis on services, she is likely to show a retardation in her rate of growth and may lose the advantage of imitation of more advanced technologies. So they caution against blindly extrapolating the data of the recent past into the future."

If you need a source check out your previous post on this topic, the Solow growth model one. Guess someone here doesn't always read the comments! No harm done.

Boonton writes:

Now I think you're confusing 'better economy' meaning a nicer place to live in with 'better economy' meaning larger gross GDP. As I pointed out on the other thread, India has a larger GDP than Australia yet no one thinks Australia should try to copy India's economic systems (granted India has been getting better over the last decade).

The Solow model does yield the possibility of the USSR catching up and even slightly pulling ahead of the US even if you assume the USSR was and always would be less efficient than the US due to their command economy.

This can happen when the less efficient economy has more labor and capital than the more efficient one. De Long demonstarated that the USSR's command economy could force consumption to be limited in order to build up capital. While in the discussion on the other thread we assumed the US and USSR equaled each other in population it appears that the USSR actually had a non-trivial edge over the US in population (http://www.foia.cia.gov/browse_docs.asp oddly I can't easily find any good site showing US.v.USSR population by year).

Long story short: Just as India is a less efficient economy than Australia but has a higher GDP do to the simple fact that it is able to throw more raw labor and even capital at production, it was a real possibility that the USSR could have 'caught up' and even surpassed the US in GDP by forcing itself to invest heavily in capital and holding a population edge.

In short 'catch up' and even surpassing were reasonable forecasts from the best available models using the best available information at the time. While your rhetorical stunt of calling the forecasts of Soviet growth "growth optimists" or "sympathetic to growth" is meant to leave the impression that pro-USSR pundits were making these forecasts, the reality is that the right was far more likely to make and support forecasts of Soviet growth.

"There was never a point in Soviet history when a sensible economist would have seen communism as good for growth in any meaningful sense."

This is a loaded phrase built on trying to tear down a straw man. No one ever asserted that the USSR's economic policies were good in and of themselves. The question that was important at the time was whether the USSR could produce growth that would support a serious military threat. The answer clearly was yes both in theory and at least with the best available data for a while (say up until the 70's when things started to grind to a halt).

Of course the USSR could have posed a bigger threat if it had adopted some less communistic policies. In that case it would have been able to threaten the US with a tripple whammy of more labor, more capital and equal or better productivity to use them. Be thankful Stalin never got a hold of China's 1990-presentday playbook from a time travelling tourist or else history could have turned out a lot nastier.

Bryan Caplan writes:

Thanks for the quote, but it only says a subset of what I said he should have said. He mentions catch-up (good), but doesn't criticize Soviet economic policies. And if he was so against "blind extrapolation," why did his projections do just that?

Boonton writes:

Do you think this is an especially fair criticism? We have a handful of cherry picked quotes and because those quotes lack "fair balance" on the USSR that's a problem?

In extended discussion on this site and Marginal Revolution it is clear no one participating in this debate actually has Samuelson's works. They have a paragraph here and there and from that they are spinning all types of fantasies.

For example, its pretty obvious to me the graph is illustrating how higher growth rates matter more over time than starting position. The Soviet Union was just a timely (and alarming) illustration of that as today China would be. A more rigerous future forecast of USSR / US GDP would have to take into account a lot more complexity...complexity which Samuelson appeared to understand in the little peaks we've gotten so far.

Anyway, this isn't a very fair criticism. If you want to say he was soft on the USSR support it with actual evidence. Because a quote here or there failed to say "and by the way the Soviet Union was evil" doesn't quite do it IMO. If I were to dig through everything you've written have you always added "by the way Hitler was evil" to all your discussions of, say, Nazi Germany? If not can I call you an unconscious Nazi supporter without your moderator jumping on my comment?

This seems to me like the right's own attempt to play the political correctness game. Being that it hasn't done any good for the left I'm not sure why you think results would be better this time around.

As for why be against 'blindly extrapolating the data'. Well there was the next part of his quote:

After that, he remarks "Although there is much that is persuasive in such arguments, there is also the fact that they may provide temptingly optimistic rationalizations to patriots who are wishful that America can stay forever ahead."

(Note that this is being pulled from a commentator on MR who said he had a copy of the 1976 edition of Economics, anyone who wants to get a more precise quote from the library w/page numbers is welcome too....sadly Google Books doesn't seem to even have a single scan of the sections in question available)

I think this very nicely shows the bind one is in with all types of forecasting. One the one hand, extrapolating the data is nice, clean and objective....plus we have lots of effective mathematical tools to do it. But it makes the assumption that the future will just look like a continuation of the past which we know rarely happens quite that way. On the other hand, trying to 'figure out' the future leaves on open to just conning yourself with a story that may sound convincing but may in fact be false.

The main thing is that Samuelson was clearly aware of the idea that the USSR was growing by copying established know how and he clearly did not believe the Soviet Union was very good at inventing its own know how. So in terms of which was better, market economies or communist ones there's no evidence to indicate he entertained the latter. The question of which country would have the bigger GDP in the future though (and by extension could send more tanks into Europe, bombers over the poles and subs into the ocean) did not require you to believe the USSR was a 'better economy' to feel there was a real danger it could catch up and even exceed the US's GDP.

As I point out in other comments simply having a bigger GDP doesn't make one country's economy better. India has a larger GDP than Australia yet no one feels India has a superior economic system.

ryan yin writes:

Boonton,
Again, the Solow production function doesn't really allow for a low total-factor-productivity ("A") economy to surpass a high TFP economy by throwing extra capital. This is precisely because it doesn't make sense for an economy with low TFP to have even as much capital per capita as a high TFP country. Your India-Australia example is kind of an "exception that proves the rule" -- total GDP and total capital are higher only because of 50 times higher population. This reasoning does not apply for countries of roughly the same population (e.g., US & USSR). (This also says that it's a mistake to distinguish between higher GDP and "better economy", or at least in the way that you do. A country with low TFP will in practice always have lower GDP per capita, which in this case means lower total GDP too.)

So even if we stay that Samuelson was right to suppose the relative gap in TFP between the two countries would not increase, and even if we assume Samuelson was right to suppose Soviet "investment" was real investment, the straight-line projections he's making are contradicted just by plain old Solow model. You're right that Samuelson had "lots of effective mathematical tools" but even those tools should tell him that Soviet growth couldn't be sustained unless their A was growing faster than the US's. It seems like we have to grant him an awful lot (and then suppose he just became really bad at math) to rationalize his projections.

baconbacon writes:

"Of course the USSR could have posed a bigger threat if it had adopted some less communistic policies. In that case it would have been able to threaten the US with a tripple whammy of more labor, more capital and equal or better productivity to use them. Be thankful Stalin never got a hold of China's 1990-presentday playbook from a time travelling tourist or else history could have turned out a lot nastier."

Really? China's playbook from the 90s has yielded them a GDP/capita of ~$3,500 a person and a PPP GDP/capita of ~$6,000- good enough for ~ 1/7th of US GDP/capita. That's with 4X the amount of labor and about a million percent better communications technology existing- with 30 years of "capitalist-communism" (or soft fascism or whatever you want to call it).

"The main thing is that Samuelson was clearly aware of the idea that the USSR was growing by copying established know how and he clearly did not believe the Soviet Union was very good at inventing its own know how. So in terms of which was better, market economies or communist ones there's no evidence to indicate he entertained the latter. The question of which country would have the bigger GDP in the future though (and by extension could send more tanks into Europe, bombers over the poles and subs into the ocean) did not require you to believe the USSR was a 'better economy' to feel there was a real danger it could catch up and even exceed the US's GDP."

This is vastly overstating the danger of a higher total GDP leading to military superiority. The total number of planes, tanks, ships, subs and soldiers produced does not determine the out come of wars. If this were remotely true the USSR would have been stuffing Germany's advance into their territory and rampaging all over German soil well before the US landed at Normandy.

Boonton writes:

Ryan

Again, the Solow production function doesn't really allow for a low total-factor-productivity ("A") economy to surpass a high TFP economy by throwing extra capital. This is precisely because it doesn't make sense for an economy with low TFP to have even as much capital per capita as a high TFP country. Your India-Australia example is kind of an "exception that proves the rule" -- total GDP and total capital are higher only because of 50 times higher population. This reasoning does not apply for countries of roughly the same population (e.g., US & USSR). (This also says that it's a mistake to distinguish between higher GDP and "better economy", or at least in the way that you do. A country with low TFP will in practice always have lower GDP per capita, which in this case means lower total GDP too.)

1. The projections we are discussing are total GDP, not GDP per capita. GDP per capita is important if you are talking about everyone having a nice home, good TV's and being able to eat out on a regular basis. Total GDP is important if you're wondering who can send the most tanks to the other side of Europe in a hurry.

2. Actually the USSR appeared to have a larger population than the US so it had an edge in the L part of the Solow function. The USSR's shortcoming was in the A and the K. Which leads to....

3. Assuming Soviet planners felt their system was equal to or better than the US why wouldn't they try to boost K? If they believed they could meet or even just keep a relatively close pace with the US's A, then it would seem the best way to beat the US would be to build up K. Once K was near US levels, the USSR would have an edge from population alone.

In other words your argument might make sense if the Chicago School of Economics combined with Warren Buffet were in charge of the USSR's output. They weren't.

So even if we stay that Samuelson was right to suppose the relative gap in TFP between the two countries would not increase, and even if we assume Samuelson was right to suppose Soviet "investment" was real investment, the straight-line projections he's making are contradicted just by plain old Solow model.

You're right that Samuelson had "lots of effective mathematical tools" but even those tools should tell him that Soviet growth couldn't be sustained unless their A was growing faster than the US's.

You're assuming he was using the Solow model. De Long's point was that the Solow Model was a neoclassical model not created or used by people with any favoritism to central planning and with reasonable assumptions it could project a catch up. Samuelson, I suspect, was simply using the 'draw a straight line' method of modeling here which lacks the calculus but is much friendlier to undergrads and quite often still holds its own compared to more sophisticated models.

Making it simple, imagine a USSR with an equal amount of K and a larger (but not huge) advantage in L.

baconbacon
This is vastly overstating the danger of a higher total GDP leading to military superiority. The total number of planes, tanks, ships, subs and soldiers produced does not determine the out come of wars.

That is very true, nonetheless I doubt any serious military expert during the Cold War would have asserted that the USSR would become less of a threat if they had more planes, tanks, bombs etc. The reason the US military clung to nuclear supremacy during the Cold War was the very reasonable fear that Soviet numbers could outmatch superior US knowhow in a full fledged conventional war. Here I think commentators mess up recent history of the Gulf War and Iraq War where the US has technological superiority many orders of magnitude greater than its enemies with the Korean and Vietnam Wars where the record was much more mixed.

Steve Roth writes:

Again, you need to look at the underlying assumption: that Europe's system has had the same goals as the American system.

I like Bernard Wasow’s quote:

"Between 1970 and 2000, GDP per person rose by 64% in the United States and by 60% in France. In America, this came about because productivity per worker rose by 38% and hours worked per worker rose by 26%. In France, it came about because productivity rose by 83% while hours worked fell by 23%."

http://www.centuryinstitute.org/list.asp?type=NC&pubid=596

Which of these describes a better approach to growth--and life?

Boonton writes:

If we use Solow's model here I think the Europe is slow and statist meme is suffering a bit.

If its true France and the US both increased GDP by about 60% between 1970 and 2000 then there are only 3 places that increase can come from. K, capital, L, labor and A 'know how' or productivity.

If America's increase was caused mostly by working more hours and less by increased productivity then that's not very promising. You're just tossing 'L' at the production function rather than K or (even better) A. There are only so many hours a person can work to begin with. This method of beating France is essentially the same 'method' India uses to beat Australia in GDP. Not very hopeful!

Likewise if France's increase was in A....well how is that possible!?

One possible answer is that 'statist' policies are good for A.

A more likely answer in my opinion is that we are reasoning by stereotypes and Fox News talking points here. How statist is France really relative to the US and vice versa? Perhaps we just think France is much more statist because we hear stories about French laws that would never be enacted in the US. But the reason we hear about these laws is because they are so odd to our ears that they make good stories.

Can someone address how exactly do we measure 'statism' here? Is France really more of a 'statist' economy? If so how much more? 25%? 100%? 1,000%?

Elvin writes:

A quick comment to Boonton:

I think the demographic factors after 1970 are very complex. You have civil rights for minorities, women's increased participation, immigration, the baby boomers entering the workforce, the aging of union workers, and more free trade. (See Samuelson-Stopler).

The net result is that I wouldn't necessarily say that productivity decreased. The quality of the labor input may have changed as well.

Some of these happened in France, too, I admit.

Boonton writes:

Lowering discrimination in order to bring women and minorities into the labor force would, I think, show up as increased labor hours.

"Quality of labor", I think, shows up as an improvement in A. For example, if a woman with a college education is able to get a high level job today whereas in the past she was only able to get a middle level position because of discrimination but still worked a 40 hour week that would show up as an improvement in 'A', not 'L'.

jowana bissaif writes:

Bryan Caplan asserts: "with better policies [i.e., less statist], Europe would still be gaining on us - or would have already caught up." In 2008, per capita personal income in Kentucky and Arkansas was $32,000, compared to $51,000 in New Jersey and Massachusetts. That's a bigger disparity than USA compared to Western Europe. The current motto of the state of Kentucky is Unbridled Spirit™. I think everyone would rate KY and AK as somewhat less statist than NJ and MA. This illustrates the very well-known fact that a whole lot of other stuff beyond state policies is in play in any jurisdiction, implying that statements like Bryan's above are indefensible oversimplifications, whose truth or falsehood can never be found out.

Jowana Bissaif writes:

Bryan says that Western European countries ought to be still be gaining ground and catching up on the US in terms of economic output per capita. I just looked at the growth data, and they are in fact still growing at a very slightly faster rate than the US. Annualized over the 18 years 1990 through 2007, the US GDP per capita growth rate was 4.17% and that was a lower growth rate than in Netherlands, Belgium, Denmark, Norway, United Kingdom, Irish Republic, Spain and Portugal. It was higher than in France, Germany, Italy, Sweden, Finland, and Switzerland, but not by much. France and Germany grew by 3.75% a year. The U.K. grew 5.82% a year, Netherlands 5.20%, Denmark 4.65%. The data is at Wikipedia under the page title "List of countries by GDP growth 1990–2007"

Doc Merlin writes:

You guys need to put Murphy in your blogroll, considering how much interaction you all have.

Comments for this entry have been closed
Return to top