Why is this “interesting?”

Over at Freakinomics Justin Wolfers makes the following observation:

There’s interesting research waiting to be done explaining just why some countries have been hit harder by the global financial crisis than others. . . .

[Graph showing bigger declines in GDP in countries with larger medium and high-skilled manufacturing sectors]

The lesson: the greater your involvement in producing high-value goods, the harder the fall. Perhaps macroeconomic stability comes despite GM, rather than because of it.

Suppose the big drop in GDP in countries like Germany, Japan and Taiwan was not caused by the financial crisis, but rather by the huge drop in worldwide aggregate demand.  In that case wouldn’t you expect output to fall faster in countries that had relative large capital goods sectors (including consumer durables like cars and flat panel TVs?)  And shouldn’t the recession be milder in countries that focus on services, such as the US and UK?  If so, then there’s really nothing “interesting” to be explained.  Sorry to be such a killjoy, but the answer is right in front of our face, we simply refuse to see it.  Monetary policy errors allowed NGDP to fall not just in the US, but almost everywhere.  The results are exactly what the textbooks say should happen.

I have no idea what would occur if the financial crisis actually was the cause of the big drop in manufacturing output, as I see no reason to think about hypotheticals that are unsupported by any evidence.  (And don’t say output fell because firms can’t get financing.  That confuses cause and effect.  These firms can’t get sales.  If the orders were flowing in, they’d find financing.)



16 Responses to “Why is this “interesting?””

  1. Gravatar of Jim Glass Jim Glass
    8. June 2009 at 13:02

    “The lesson: the greater your involvement in producing high-value goods, the harder the fall.”

    I agree with you.

    Moreover, Wolfers confuses “high tech” with “high value”.

    Back in his pre-NY Times days, when Krugman used to bash the left, he regularly riduculed guys like like Thurow and Reich who proposed national industrial policy initiatives on the premise that “high tech = high value”.

    Many, many, high value items have no “high tech” to them, and high-tech products tend to rapidly fall in price (and value) as they become commodities. (Due to Moore’s Law and all, think of last year’s chips and circuit boards that are this year’s big commodity sellers.)

    It may well be that economies that depend on “high tech” exports are suffering as world demand for commodities such as circuit boards, dvds, etc., plunge.

    That’s rather different from saying economies with “high value” exports are suffering.

    Though Bryan Caplan also points out that judging solely from the graph it’s equally as plausible that it’s just that those four Asian economies are suffering disproportionately for a regional reason of their own. Drop them off the chart and that line looks pretty much vertical, regardless of the “value” of the nations’ exports.

  2. Gravatar of Alex Golubev Alex Golubev
    8. June 2009 at 15:14

    what i find even more interesting is that there’s a possibility that the cause and effect of financing vs decision to make a trade are even arguable. Isn’t that an argument that our models should reflect reinforcing processes? I don’t really want to argue what you’re suggesting with relation to the graph, but it’s hard to argue that people ignore the cost of capital in general or that underwriting guidelines don’t get tightened when a bank’s balance sheet is under duress even from an UNRELATED asset class. Why no feedback loops/differential equations? why just, “A caused B period?”. I don’t even want to pick on economics that much. this is evident in all social sciences as well. when uninteligent physcial processes have reinforcing processes, how can social phenomena NOT possess these qualities?

  3. Gravatar of Nick Rowe Nick Rowe
    8. June 2009 at 17:56

    But why did so many central banks around the world make the same mistake at the same time?

    Not all countries have fixed exchange rates with the US$. In principle, it should have been possible for (say) the Bank of Canada to have loosened monetary policy, and kept AD (or NGDP if you like) in Canada from falling, even if the Fed kept monetary policy too tight.

    I have been puzzling over this question for some time. I have 4 possible answers, but none of them is satisfactory:

    1. Events moved too quickly, so they were all behind the curve.

    2. They all had the same theoretical perspective on monetary policy, so all made the same mistake in responding to a common shock.

    3. If one or two big countries get monetary policy wrong, that causes a real shock to the rest of the world (AS in ROW shifts left), and monetary policy cannot fix AS shocks.

    4. Our theories of AD determination in open economies are wrong. It is not possible for one country to use monetary policy to keep AD (NGDP) growing on target if all other countries get monetary policy wrong, even under flexible exchange rates.

  4. Gravatar of Jon Jon
    8. June 2009 at 18:28

    Nick: Great list! I’ve been thinking along the lines of #2-#4.

    I think #3 is the most conservative, but #4 interests me quite a bit intellectually.

    In particular, I have an intuition that indeed much of the edifice of modern theory applies only to a closed economy. Interest-rate arbitrage dominates currency values and drives liquidity out of the country of origin. This is a triple upset of conventional theory. Local effects are attenuated, global effects are accentuated, and the exchange-rate shift induces import-inflation.

  5. Gravatar of Richard A. Richard A.
    8. June 2009 at 20:03

    A possible explanation is that world wide too much attention is paid to price inflation at the expense of nominal GDP growth. The oil price shock may have caused the central banks to under grow their countries money supply at the expense of nominal GDP.

    This would fit with Nick’s #2 explanation.

  6. Gravatar of Bill Woolsey Bill Woolsey
    9. June 2009 at 03:31

    Central banks worry too much about exchange rates? Central banks worry about something like the Consumer Price Index, and include the prices of imported consumer goods in the calculations? Central banks believe they must avoid competitive devaluations? Central banks worry that the U.S. will retaliate against exchange rate deprecisation with tariffs and quotas that will persist after the crisis? Central bank worry that U.S. investors in their own debt will be deterred from future investments by an exchange rate depreciation?

  7. Gravatar of ssumner ssumner
    9. June 2009 at 04:46

    Jim Glass, Yes, I agree that “high tech” is very misleading. The “soft” part of Dell’s business (design, marketing, customer support) is arguable harder to do right than the “hard” part (component manufacturing.) Otherwise more of the value added would go to countries like Taiwan.

    I saw Bryan Caplan’s post,


    but I am not willing to dismiss the Asian countries. The sub-prime crisis was concentrated in the West. The biggest Asian market (in PPP terms) is still growing. The Asian exporters are getting killed by the drop in western demand for manufactured goods. Manufacturing has always been more cyclical than services. Bryan’s right that the graph doesn’t show a particularly strong relationship, but I still think Wolfers is on to something. Another interesting case is Germany, which didn’t have much of a housing bubble, and initially felt schadenfreude during the US subprime fiasco, but now is suffering worse than we are.

    Alex, Yes, there are lots of feedback loops. But my main point is that there is no “interesting” problem to solve—manufacturing is very cyclical.

    Nick, I answered a similar question in my newest post before I read this. I basically took the position that #2 and #3 were the answer. I can’t buy number one as an excuse because yield spreads showing inflation expectations are observable in real time. And at worst they should have immediately moved rates to zero in early October. Yet the ECB’s rates were still 4.25%, and even today haven’t yet been cut to zero. Apparently the ECB expects on target nominal growth, which makes you wonder what their target is.

    On point four, I strongly disagree with the view that small open economies cannot boost their nominal GDPs in a recessionary world environment. Zimbabwe successfully swam against the tide, so why can’t Canada? [Now Bill’s going to scold me for feeding the illusion that NGDP targeting is a code word for inflation. So let me put it this way: If NGDP grew 5 billion percent in Zimbabwe, then Canada could do 1/1,000,000,000 as much stimulus as Zimbabwe.]
    But I am not sure they want to target NGDP. And if they do, they may still see a sharp fall in RGDP due to falling world AD and hence falling demand for Canadian exports.

    BTW, I just read that Latvia may have to devalue. If they do then there is a good chance that I will have been wrong in my support of Estonia’s currency board, and Bill will have been right.

    Jon, See my response to point 4 of Nick.

    Richard A. Good point, but didn’t you mean to say “contract their money supply?” But yes, focusing on the CPI rather than NGDP last year was a huge mistake.

    Bill, Take away those question marks—those are all good explanations.

  8. Gravatar of Jon Jon
    9. June 2009 at 05:51


    I’d be willing to concede that NGDP could be grown Zimbabwae style–but you still haven’t convinced me that a NGDP target wouldn’t devolve into negative real growth and lots of inflation.

    I always thought this was the lesson of the 70s.

  9. Gravatar of Nick Rowe Nick Rowe
    9. June 2009 at 05:52

    Scott: The Zimbabwe counter-example makes sense. Though you could argue that Zimbabwe was different, because it didn’t really have a financial sector to begin with.

    I’m leaning towards 2 with elements of 1. They had a common theoretical framework that ignored the early-warning signs from asset markets, so got behind the curve.

    (I think you mis-read Richard A. He said “under grow” the money supply. Same as “contract relative to what it should have been”.).

  10. Gravatar of ssumner ssumner
    10. June 2009 at 05:58

    Jon, That was not the lesson of the 1970s. In the 1970s RGDP growth was fairly normal. The problem was that NGDP growth was extremely high, and thus so was inflation. With 5% NGDP growth the 1970s would have had the same RGDP growth (just under 3%) but much lower inflation.

    Nick, I suppose it depends on the definition of “financial sector.” I presume they have banks, but I have to admit that I know almost nothing about their economy, other than that the LRAS has been moving left.

    I think we are pretty close in our view of what went wrong with monetary policy.

    Yes, you are right that I misread Richard.

  11. Gravatar of Current Current
    10. June 2009 at 10:40

    The interesting thing is that it backs up the case for the Austrian Capital theory.

    There are two ways of looking at it. Some say “services” are not very cyclic and “manufacturing” is.

    Another would be that production of higher-order goods with specialized capital equipment and knowledge is very cyclical. This latter idea indicates that Austrian Capital and business cycle theory contains a fundamental truth.

  12. Gravatar of ssumner ssumner
    11. June 2009 at 05:03

    Current, I’m not sure exactly what you mean by “fundamental truth.” If you mean that Austrain BCT explains an empirical regularity of the business cycle, that investment is especially cyclical, then I agree. But other theories such as the Permanent Income Theory can also explain that. Indeed I think all business cycle theories predict that investment will be especially cyclical. So it doesn’t show that the ABCT is the right way to think about it. Although of course it may be.

  13. Gravatar of Current Current
    11. June 2009 at 10:56

    The permanent income theory could explain it. The permanent income theory could be grafted onto either homogeneous capital or heterogeneous austrian capital. The same goes for external shocks and collapses in aggregate demand, as far as I understand it.

    The point I’m making though is mostly about capital theory not business cycle theory.

    Say we have a homogeneous capital K. If something happens to change it’s price then it should happen across that capital. If the net present value of one sort of capital falls so should all others, ceteris paribus. If that doesn’t happen the capital isn’t homogeneous.

    As far as I can see in this model two things can cause particular businesses to lose value. Firstly the behaviour of customers preferences and secondly contract customers have signed.

    Now, if manufacturing is failing because LCD TVs are optional purchases that isn’t interesting. Similarly, if it is failing because customers have signed contracts that isn’t interesting either.

    However, if it is failing because it’s capital is not substitutable then that *is* interesting in my view.

    As a Brit I can’t help wondering why the UK isn’t suffering so much. Financial services such as floatations and buyouts are often optional purchases and not based on regular contracts.

    To me the Austrian view makes much more sense. I work in the “soft” part of a high tech business. If I were no longer needed the capital equipment I use would be idle. A million dollars or so would go down the drain.

  14. Gravatar of ssumner ssumner
    12. June 2009 at 05:13

    Current, Are you saying that even though the UK is much more service-oriented than Germany, they are services that out to be as highly cyclical as capital goods? (financial services, software, etc.) If so, it is a good question that I don’t have an answer for. I just don’t know enough about that part of economics, nor am I an expert on the UK economy.

    Is software less cyclical than hardware?

  15. Gravatar of Current Current
    15. June 2009 at 03:17

    ssumner: “Current, Are you saying that even though the UK is much more service-oriented than Germany, they are services that out to be as highly cyclical as capital goods? (financial services, software, etc.)”

    Yes, exactly.

    ssumner: “just don’t know enough about that part of economics, nor am I an expert on the UK economy.”

    Well, neither do I. I’m just pointing out that what we are seeing may be interesting.

    ssumner: “Is software less cyclical than hardware?”

    I don’t think it is. Well, in my personal experience both are very cyclical.

    (From the point of view of Austrian Capital theory software is various. Take a company like Oracle for example. Their capital value is tied up in their products and processes. The codebase of their database programs are capital equipment and are very valuable.

    Other sorts of software business are different. A company that writes small websites the use databases for example. Such companies generally own little unique software of their own. They make things from existing toolkits. They have little capital that is specialized and sunk.

    So, Austrian Capital theory would not predict that the software industry would behave in one specific way.)

  16. Gravatar of ssumner ssumner
    15. June 2009 at 05:29

    Current, I think we are both reaching the limits of what we know here. Maybe someone else knows whether software is more cyclical than hardware. In any case it might be better to move the discussion to the newer post on the UK, where more people with knowledge may contribute.

Leave a Reply