Wouldn't you know it, but no sooner had Eugene Volokh of the Volokh Conspiracy linked to Impearls' table of the “largest economies of the world,” (which might be titled as above, by the way) — and just after a couple of interesting comments arrived — our Internet connection/phone line went down for two days!

In the meantime, Eugene posted a couple of updates to his initial posting, here and here (or just scroll up).  In his first post, Volokh wrote:

Impearls has a really cool table, drawn from multiple sources; I can't vouch for the accuracy, but it seems right, and very interesting.  It also points out some interesting controversies about the actual sizes of various economies, especially China's and India's — the CIA Factbook seems to give a different result from the one that Impearls says is generally accepted by economists.

Oh, and according to one source that Impearls cites as being quite reliable, Los Angeles County's GDP is right above the 16th largest national GDP — which is Russia, the main heir of the nation from which my family moved to L.A. County.

Russia is another of those countries whose CIA Factbook-stated GDP differs substantially (by a factor of four in this case) from that generally used by economists.  As we now know (see below), this is due to the “Purchasing Power Parity” [PPP] index that the CIA uses rather than exchange rates for its GDP estimations.  According to the CIA Factbook PPP-derived figure, Russia's GDP ($1.2 trillion US$) actually lies up near that of California (taken as a whole) or countries like Brazil and Italy.  Taking LAEDC's (Los Angeles Economic Development Corporation) more conventional (exchange rate) GDP figure for Russia, on the other hand ($310 billion US$), Russia's economic output hovers near that of nations on the scale of Taiwan and the Netherlands — nations small in geographical extent, and relatively small in population as well.

The latter GDP figure for Russia, in fact, lies on a par with U.S. states Massachusetts and New Jersey.  It's worthy of note, I think, that these two American states are just about the same size (around 20,000 km²) as what has been called “that sh--ty little country” (sorry! I disavow it!), Israel.  Massachusetts, in particular, is a virtual twin of Israel in area and population.  (Israel itself doesn't show up in the top economies table as its GDP — by the Factbook, $119 billion US$ in 2001 — places it below the table's $200 billion cut-off.)

My point, which just reinforces what Eugene was saying above, is that a country need not be large in area, nor very large in population, to have a vibrant, influential economy; and even without a productive economy, a small country can potentially (especially in these days of weapons of mass destruction) have an effective and dangerous military.  Thus, with an (exchange rate) economy no larger than that of the Netherlands or the state of New Jersey, Russia still fields thousands of nuclear weapons.  Iraq, with a much smaller economy, still constricted by UN sanctions ($59 billion US$ in 2001, according to the CIA Factbook), is attempting to do less, but still enough to potentially kill millions of human beings.  Note that North Korea manages an effective WMD program with an economy ($21.8 billion US$ in 2001) less than half the size of Iraq's present one.

Some of the impetus behind thoughtless and bigoted comments like the “sh--ty little” one results, I think, from a well known phenomenon in mental affairs, “Out of sight, out of mind.”  Countries the size of Israel are almost lost (say) on your average desktop globe; the tiny scale constricts visibility, and people tend to evict unseeable things from their minds as too trivial to be of concern.  This is a severe conceptualizing error.  “Small” nations (or states) need not be small in their consequences for the present or future.  (As an example, see Alexis de Tocqueville on the origins of American democracy in the little state of Connecticut and its neighbors, here.)

Interesting comments have come in from readers following the link from the Volokh Conspiracy.  I'd suspected the Conspiracy's high-powered readership would give my problem short shrift once it chanced being brought to their attention!

A writer from the U.S. Census Bureau (who also cc'ed Eugene in his reply) clears up the mystery:

Interesting post about international GDPs.  I thought I'd email to let you know that the difference between the CIA's numbers and LAEDC's isn't really so mysterious.

The reason the CIA figures differ from the LAEDC's, is that the CIA is converting from foreign currency to dollars using a “Purchasing Power Parity” (PPP) index, while the LAEDC is using exchange rates.  That's also the reason why the LAEDC and the CIA report the same numbers for the U.S.: because there's no need to convert U.S. GDP from a foreign currency into dollars.

For most purposes, the PPP numbers are preferable, so the CIA is reporting better numbers.  Two problems with exchange rates are that they are set by government fiat in some countries, and that they only reflect the price of tradeable goods.  But most of GDP isn't tradeable, for example, housing and cardiac surgery.  PPP numbers are an attempt to calculate the price of all goods in a country relative to U.S. prices.  If your goal is to use GDP per capita to compare the standard of living in different countries, the PPP numbers are the ones to use.  The exchange rate GDP numbers tell you how much the country could import (assuming the exchange rate is set by the market, anyway), but that's not a figure that I see any particular use for.

If you want to see the PPP and exchange rate numbers side by side on the same page, the World Bank has a table on the web:  [here].

Another reader commented along similar lines, but is worth quoting too for his slightly different information/slant on things:

Well, the explanation for the difference is that the CIA uses PPP-adjusted figures, while the OECD figures are based on official currency exchange rates.  In countries with relatively free movements of capital and goods, these would mostly be the same, but in many developing countries there is a substantial difference.  (So, for example, food and housing is a lot cheaper in dollar terms in Dehli than in LA.)  In places with substantial differences, there is likely to be a big black market in foreign currency.

PPP is Purchasing Power Parity, that is, adjusting the exchange rates according to what good actually cost in the country.

Which one matters more?  Hard to say.  If you're living in the country and looking at quality of life, then PPP matters more, since that mostly determines your standard of living.  If you're investing in a country, building a factory, exporting, or importing (legally), etc., then exchange rate matters, since that's what you pay and get paid to get goods and capital and equipment in and out.

The ratio of the PPP and exchange rate figures is usually about the exchange rate of the black market.

Once you look at OECD's PPP-adjusted figures, there is no discrepancy, so the CIA isn't measuring any underground economy (which by definition isn't measured).

CIA's explanation: [here] (see section on GDP methodology).

OECD figures: [here] (Compare GDP figures, PPP and exchange rate for Mexico, for instance).

I find this encouraging.  Since noticing the discrepancy between the CIA's and economists' GDPs, I'd carried something of a cloud in my mind about the reliability/usability of the CIA's GDP figures.  Now that we understand where it's coming from, they do seem to be relatively solid and usable data.  (The World Bank table cited by the first reader above is also very much worth locating; I'd been looking for a table like that!)

Ultimately, I agree with Eugene — this sort of table does provide a very interesting vehicle for comparison between the states and nations of the world.  Who says economics is the “dismal science”?