Sumner is Now More Wrong
by Mike Kimel
Sumner is Now More Wrong
So there’s a a response from Sumner to my critique of his earlier post. And he clearly doesn’t get it. Again, I’ll focus on the low hanging fruit, but this time I’ll try to simplify a bit more.
First comment here:
Um, no you don’t want to adjust for inflation when comparing two countries at the same point in time, you want to adjust for PPP. Time for Angry Bear to go back to the Ivory Tower.
This by itself has a bunch of errors.
1. I agree that adjusting for inflation is un-necessary. That doesn’t mean you don’t want to do it. I picked the first measure I came upon that looked reasonable. Inflation adjusting in a single currency and PPP generally produce the same results. Apparently they aren’t doing it for Sweden to US comparisons for the years he picked. Very odd.
2. I am not Angry Bear. That’s the name of the blog. Several people post regularly at Angry Bear. Rdan, the site owner, is kind enough to put up my posts when I write one. But calling me Angry Bear is like me calling The MoneyIllusion.
3. To which Ivory Tower would he suggest I return? Last I checked, Sumner is the one whose bio says something about spending the last 27 years teaching at Bentley. Nothing wrong with that, of course. And for what its worth, I did the adjunct for five years too, which is the closest I’ve come to being an Ivory Tower guy. But I’ve been in the private sector – working for a Big 4 firm, two Fortune 500 companies, and as a consultant under my own shingle – since leaving grad school. So please, spare me.
Next he tells us how he picked his sample:
I didn’t have time to take all 200 countries, so I did what I thought would be interesting examples.
OK then. There’s a fine line between that and cherry picking unless we’re told more, and we aren’t being told more. Now, that doesn’t mean he isn’t cherry picking the data, but this is one heck of a “trust me” given what he’s trying to sell.
There’s a lot more here, but I’m going to focus on what I mean by what I wrote when I said Sumner clearly doesn’t get it. Let’s take this for instance:
Argentina ran statist policies in the 1980s and 2000s. There were reforms in the early 1990s, but free market policies don’t produce success when combined with contractionary monetary policy. Indeed that’s essentially the whole point of my blog. Argentina did very well for a while after it did some privatization, but of course their ultra-tight monetary policy put them into a depression at the end of the 1990s. Then they switched back to statism. I have no idea why a country with statist policies in the 1980s and 2000s and a deflationary monetary policy in the late 1990s, would be expected to do well.
As he said, that was the post of his earlier post. So its a good place to focus. But before I get started, let me warm up with a quibble…
Argentina was not doing well in the early to mid-90s. That was a, well, a Money Illusion. See, they sold off all the state owned enterprises (and the strangely enjoyable to look at water building ) and then lived high on the hog off the money. But living off a one-time chunk of revenue as if its a long term revenue stream may make you look impressive to folks who think that reading a newsmagazine qualifies as knowing how the world works, and it even fooled a lot of folks in Argentina, but it didn’t fool everyone even if most people in Argentina were hoping for the best. And it didn’t fool reality, and the money soon ran out. It always amazes me that most people who call themselves economists and mention Argentina haven’t figured out that detail.
And now to the meat (sadly, not Argentine) and potatoes… which can be summed up with this sentence: “I have no idea why a country with statist policies in the 1980s and 2000s and a deflationary monetary policy in the late 1990s, would be expected to do well.”
And yet, he bats not an eye in his first post when he points out that Japan and the US grew about the same. But Japan is probably more statist (think the Ministries) now than privatized Argentina, and the trend away from statism was certainly greater in Argentina during the 1980 to 2008 period than in Japan. As to deflationary monetary policy – exactly how long has Japan’s interest been about zero? So shouldn’t Japan count as a loss in Sumner’s book?
But that’s trivial. A bigger problem is that Argentina’s deflationary monetary policy could be summed up in one word: dollarization. The currency was pegged to the dollar, and the rest was a collection of details (some of which made life miserable for the average person in Argentina).
Now, regardless of the reasons given, the effect of a peg is the same. And it turns out that among the success stories on Sumner’s list was Hong Kong, which has been operating with a similar currency peg for a lot longer than Argentina ever did. (BTW, the Hong Kong government had a cool little study looking at their peg and Argentina’s here.)
I’m not all that familiar with Singapore, but I do know the Singaporean central bank is kind of an odd duck in that they use monetary policy to regulate the exchange rate (which is a fancy way of saying they peg the currency, but move the peg). Which is another way of saying that on the big issue, they follow the same monetary policy that Sumner thinks is a no-no with Argentina. And of course, if you had to pick which of two countries, Argentina which sold off its state owned assets, and Singapore which didn’t, liberalized more in the last three decades its a no-brainer. Singapore, complete with its canings, has a more functional rule-of-law, but in his post, Sumner talked about how ” neoliberal reforms after 1980 helped growth.” Argentina had ’em. I don’t know about Singapore. But he has Singapore as a success story, and Argentina not.
Moving on (this is getting to be a pain in the neck) we have Chile, the other Latin American country on his list… and like Hong Kong and Singapore, also a success story. Now, both Argentina and Chile chucked their military overlords between 1980 and 2008, but the Chicago Boys like to point out they were doing their thing during the Pinochet years too. Put another way, whatever liberalization happened in Chile post 1980 (off the top of my head, I’m thinking mostly a big change to mining law in the 90s… but note that CODELCO, the state owned copper firm, still runs the big show) pales in comparison with Argentina selling off its state owned assets during the same period.
Which leaves us with three of Sumner’s four success stories (Hong Kong, Singapore, and Chile) at odds with his point. And that’s out of 13 observations picked entirely not at random. What can I say – it seems more than a bit odd.
It appears this is not a normal distribution.
Actually, you do have that wrong. Inflation adjusting doesn’t account for changes in the exchange rate, which is what PPP is supposed to correct for.
If Mexico’s GDP per-capita is 1000 pesos and the US’s is $1000 while trading at 10 pesos to the dollar, then in dollars the US GDP per-capita is “10 times” Mexico’s. If the exchange rate were to suddenly fall to 20 pesos to the dollar, then US GDP per-capita jumps to 20 times Mexico’s, with no time for inflation to do anything at all.
You can look at the ratio of real GDPs each in local currencies, or you can use common international dollars (current, or real– the latter because you’d use the *same* deflator for each series, so it cancels) But to use the ratio of real exchange-rate GDPs is not actually kosher.
Hi, I’m the david that commented a bunch on Sumner’s response post.
Singapore indeed “neoliberalised” across the 80s and 90s; it did so by splitting (already fairly efficient) state-run businesses into multiple partially private corporations all majority-owned by the government’s two investment arms.
Whether or not this counts as privatisation is up for debate, I suppose. The explicit aim was to encourage competition among government subsidiaries, not to surrender government control; this philosophy is why Singapore now has two bus services, (until recently) two taxi companies, (until recently) two media broadcasters, two investment arms, two port operators… this even extends to non-profit government-linked organisations like science research (2 eye research centers!) or whatever. Just how much meaningful competition all this induced isn’t clear. Singapore is more tightly disciplined by international markets than any domestic initiatives anyway.
It’s worth remembering that Singapore doesn’t just own most of its businesses, it also owns its organized labor (Singapore is highly unionised, but the National Trades Union Congress is a docile party organisation, and the government works very hard to collapse the state, the city, the government, and the ruling party into one entity). This is a relic of its socialist past, but if you squint the right way you could make the argument that it’s formally controlled by its organized labor. Sort of.
Let’s just say that Singapore is an unusual case.
Notwithstanding your response to me yesterday (which I appreciated), as these back and forth rebuttals get longer and longer (not to even mention david’s quite interesting points above about Singapore), I believe more and more firmly in my original point.
Almost all conclusions are illusions. Maybe we all might get along a little better if we didn’t constantly overestimate our ability to understand and make sense of a very, very complex world.
Note that this DOES count as Hayekian — the wiping boy of leftist hacks who write about “neoliberalism” without knowing what they are talking about.
David writes:
“Singapore indeed “neoliberalised” across the 80s and 90s; it did so by splitting (already fairly efficient) state-run businesses into multiple partially private corporations all majority-owned by the government’s two investment arms.
Whether or not this counts as privatisation is up for debate, I suppose”
Sumner’s shortomings on this topic come mostly from not knowing enough (or researching for the post) about each country mentioned and from not having an entirely clear definition of neoliberalism.
Your own post reveals you have a more extreme version of these same shortcomings.
Out of curiosity… in what way is it Hayekian?
sounds like you’re pointing out some cognitive dissonance
run,
I agree. But I’m more disturbed by the fact that the observations seem to be picked to show something. (And worse, after that, they don’t.)
D R,
I do precisely zero cross-country in the real world, and this may be my first post looking at cross-country data. So I may be well out of my depth… which is why I kept to Sumner’s approach, which is discuss a few anecdotes. (I picked up enough about Argentina through osmosis to spot the problem in his first post immediately, and as he pointed out in this one, its his key point.)
All that is to say, I am not an expert on PPP. But… (and I may well be wrong), your example depends on how often GDP is calculated, how often the exchange rate is checked, and how often it is PPP’d, so to speak. And my assumption – and again, I may be wrong – is that all this is done for most of these countries once a year, or four times a year at most.
My understanding is that PPP may, in theory, may be a better measure, but then, in theory, the market exchange rate should converge that way anyway (particularly if most of the country’s goods are tradeable). In practice, its an additional layer of guesswork, with the need to come up with “baskets” of goods and services that are viewed similarly in each country.
But this is a problem. No two countries do value goods similarly. And then there are quality issues. I’ll take an Argentine steak over an American one any time. Or dulce de leche over caramel. And caramel could disappear in the US completely and who’d notice, but dulce de leche is part of the culture in Argentina. Hence, compound the guesswork and approximations that go into GDP calculations with additional guesswork and approximations and you have the PPP adjusted GDP. Thus, when the first figure I came up was GDP adjusted into 2000 dollars, and being in a hurry, it seemed a good place to stop.
(Aside: When I got married, my wife and I took a trip down to Argentina so she could meet my relatives. While there, for grins, so she would know that even familiar things could be very different, I told her we should walk into a McDonald’s. (If anyone reading is familiar with the area, we walked into the one a stone’s throw from the Recoleta cemetary.) She felt a need to a) take a bunch of pictures, and as I recall, put them on her facebook account and b) try some of the gourmet items on the cafe menu.)
david,
One of the nice things about a blog like this is that you learn more from reading the comments to the posts. I read your comments on Sumner’s post, and they’ve been educational. (My knowledge of Singapore was limited to knowing about canings, no gum chewing, and an oddly behaving Central bank.) It sounds like an interesting place which I wouldn’t to visit in the summer.
AJ,
Some things do stand the test of time. Some don’t. The world is complex, but if we don’t try to understand it, our decisions will make us poorer.
Greg Ransom,
And Hayek (at least the Road to Serfdom period Hayek), no doubt, would have been even more thrilled that Argentina privatized its state-owned enterprises. And he would have ecstatic with the deregulation in the financial sector here in the US.
Hard and fast rules of thumb (whether on the left or right or libertarian wing) that don’t conform to the cultural norms of a how country operates can lead to some really crazy outcomes.
James,
Please point out examples where I have these shortcomings so I can correct them. Otherwise this just sounds like trolling.
Jennson,
It just seems to me that he isn’t willing to accept facts that contradict his views. I mean, it wasn’t that long ago that Argentina was the poster child for folks who believe in the approach that Sumner seems to hold. The fact that the outcomes were awful is the only thing that changed, but its enough to pretend that Argentina was not following that approach. You can’t lose this way.
The biggest problem with Sumner’s post is that it’s anecdotal. As you say, choose a bunch of countries, then tell stories about them. A freshman in college could have written his post. He doesn’t show any statistical correlation (ivory tower stuff, I know).
Without correlation, it’s damned hard to assert causation…
If government size as a % of GDP is a reasonable proxy for “statism,” Sumner’s correlation doesn’t exist (for prosperous countries over the last several decades):
http://www.asymptosis.com/europe-vs-us-who%e2%80%99s-winning.html
http://www.asymptosis.com/small-government-spurs-growth-economists-say-no.html
http://www.asymptosis.com/an-open-letter-to-robert-barro.html
Well, he’s got another one up. I guess I’ll write a response tonight. But I see the problem.
When a government minister tells companies where to put their R&D money and the companies do it, that strikes me as statism. When the government owns 60% of the economy, that also strikes me as statism.
On the other hand, I look at a country where the idea that someone might have listened to the government at some point about something is treated as an urban legend and I conclude the government doesn’t have much of a say in running this economy.
Clearly Sumner’s views on this basic principle are the opposite of mine.
All that needs to be measured is growth in real GDP per-capita measured in local currency. Taking the ratio masks which country is moving. PPP is useful for estimating the relative size of the economies, but not necessarily a good measure of one country’s growth.
There is nothing wrong with simply saying country X grew 5% in real (2000) dollars per capita while country Y grew 7% in real (2005) euros per capita. Nor is there anything wrong with looking at the /change/ of the ratio of the two even if the level isn’t meaningful.
If country X went from 1000 to 1100 year-2000 dollars and country Y went from 1,000,000 to 1,210,000 year-2005 pesos, then the ratio of Y to X goes from 1000 to 1100, indicating Y grew faster than X. It just doesn’t imply anything about Y being 1000 times more productive.
As far as the PPP stuff goes, you are correct that there are survey method and frequency issues. In fact, between surveys, the PPP GDP is probably computed from local currency GDP anyway– this can be confirmed by looking at the two measures using the IMF’s WEO databases. Growth in real PPP GDP and growth in real local currency GDP are almost identical almost everywhere.
…Which is why it’s best to use real local currency instead.
“On the other hand, I look at a country where the idea that someone might have listened to the government at some point about something is treated as an urban legend and I conclude the government doesn’t have much of a say in running this economy. “
It isn’t weather or not people listen to the government that makes a country economically statist. If it was you would have to rank the soviet union as a fiscal libertarian paradise because the average person in their day to day economic affairs ignored the government and used the black market.
What makes a government economically statist or not is its willingness and ability to exert coercion on their own people for economic ends. By this metric Argentina is very statist, but Singapore isn’t. Singapore is very statist in other ways, but when it comes to economics they tend to be more free market than most latin American countries.
Also, cut back on the snark. It doesn’t promote good debate.