« Tree Owners are Tree Huggers | Main | Can we do without digital rights management? »
Purchasing power parity
1. A tautology.
2. An equilibrium condition, satisfied by any good theory of exchange rates.
3. Itself a theory of exchange rates.
4. An outer bound which good theories of exchange rates may not violate.
5. Holds only for individual goods of homogeneous nature.
6. Holds only for price indices of tradeables.
7. Totally false.
8. All of the above.
Esoceicy! Correct answers will be given a free dining room set.
Posted by Tyler Cowen on February 12, 2007 at 07:49 AM in Economics | Permalink
Comments
The inspiration for the best table in the Economist.
Posted by: nelsonal at Feb 12, 2007 8:23:19 AM
That's a terrible prize. The shipping costs would kill me. And it probably wouldn't be in a style I would like. Could you just sell it locally and send me what you get? Even if it's only ten bucks I'd probably be ahead.
Posted by: Jason at Feb 12, 2007 8:30:28 AM
3 and 6 are the closest I think. I would also accept 7 because I think true PPP takes into account the local level of income. That's why it is cheaper (at current exchange rates) to buy a good meal in Anderson, SC than it is in New York City, and cheaper still in Poland or India.
Posted by: Mike N at Feb 12, 2007 8:35:53 AM
Given that PPP famously fails even the simplest of tests (witness the famous paper about various IKEA goods from a few years back), we ought add
#9: Non-tariff market barriers, non-tradable components, and regional market power imply that PPP is unlikely to hold anywhere.
As an example, have the prices of American exports sold in Europe dropped by the 30% or so PPP would suggest over the last few years?
(And further, are there any forex parities other than CIRP that hold? UIRP isn't even close to evident in the data, for example, and there's much more reason to suggest that should hold versus PPP.)
Posted by: cure at Feb 12, 2007 9:19:02 AM
Tyler, I won a dining room set when I was on "Jeopardy" years ago. I had no room for it, no particular liking for its style, and little hope of being able to store it and subsequently sell it. And, naturally, I would owe taxes on it in the meantime. I ended up turning the prize down, as I think a lot of people do. Don't inflict another one on me. . .
Posted by: Derek Lowe at Feb 12, 2007 9:25:19 AM
My girlfriend just purchased a dining room set for us, as an early birthday gift for yours truly. We received a "free" rug with it. I've no idea what I'd do with this second dining set, perhaps light it on fire.
As to PPP, I have to agree with Mike N that #3 and #6 seem closest. However, I don't find #7 acceptable, because I seem to recall (from the deep, hazy recesses of memory that store the few classes I took on international finance) that PPP is a very long-run phenomenon, so if you average out exchange rates over a long enough period they'll tend toward PPP...so I don't think it's totally false.
Posted by: Timothy at Feb 12, 2007 9:50:06 AM
Alex sent me a query, here was my response:
I've been waiting for someone to ask...
"Esoceicy n. Obsession or preoccupation with the little things."
Of course poking fun at the notion that a) to most people it does seem like a little thing, b) it really isn't little at all, hardly anyone understands a very fundamental theory!
But that is only in an on-line Dictionary of Nonsense Words, it doesn't seem to be a real word at all, which heightens the Borges-like use of the term and the reference to the unreality of PPP. There are other aspects of the reference, too.
But of course only the most loyal of MR readers will get those jokes.
Posted by: Tyler Cowen at Feb 12, 2007 10:21:27 AM
i would say that, in absence of trade barriers, it would be 2 - meaning that in a simple model of an open economy, where we have only tradable goods and no taxes and tarriffs and no transport costs, the absence of PPP is equivalent to an arbitrage opportunity; we would trade on this arbitrage opportunity as any rational agents would, until it disappears.
I would not say that it is satisfied by "any good theory of exchange rates", since a good theory of exchange rates should be more complex. However, if simplifying my model would lead to a special case where PPP does not hold, i would say "ooops!" big time.
I do not know if indeed, it holds for indexes, but it does not hold for homogenous goods (if nothing else, the Big Mac index is there to prove it).
My guess is that it simply does not hold, at least not on the short term - simply because exchange rates nowadays are not determined by goods trading, but by asset trading (i.e. we can invest abroad, we can even consider investing in a currency, so why should an exchange rate condition specified solely in terms of trading goods hold?).
The reason why i chose 2 over 3 is the following - PPP is a condition derived from a set of assumptions, hence, I feel uneasy to call it "a theory", but if i give wording another thought, i may change my mind. hmmmmm
Posted by: avm at Feb 12, 2007 10:24:24 AM
1,2 and 3 all seem correct to me.
Posted by: josh at Feb 12, 2007 10:49:45 AM
Esoceicy is a perfectly cromulent word.
Re: PPP, shouldn't we see the differences in prices of nontradeables as an implicit tax or lifestyle charge? When you buy a Big Mac in Manhattan or in Paris you're also paying for the cheap, accessible cultural opportunities; for the proximity to financial centers; and for unionized government employees. You don't get equivalent benefits when you buy cheaper Big Macs in the Deep South, let alone less developed countries.
I don't see any reason why such implicit taxes should equalize across countries in the long term, unless we expect incomes and preferences to equalize as well.
Posted by: DK at Feb 12, 2007 10:55:40 AM
Well, if the offer is inclusive of shipping costs, then the answer is no:2.
Posted by: GVV at Feb 12, 2007 11:04:21 AM
hence the tautology, DK.
Posted by: josh at Feb 12, 2007 11:05:23 AM
Esoceicy embiggens even the smallest man or woman.
Probably 3 & 7 but I'm not an expert.
Posted by: eriks at Feb 12, 2007 11:44:38 AM
Recalling Isard's critique of an economic theory that projects all economic activity on top of a pinhead - even if we think of theories of exchange rates -, it is not hard to verify that a good such as a hamburger can hardly be thought to be homogeneous if it is sold at different locations, which necessarily implies that PPP need not hold for that sort of good.
Without transaction costs of any sort, 5) should definitely be true if we can think about goods that are homogeneous in the economic rather than in a physical sense.
I think a good theory of exchange rates may not violate this fundamental theoretical result, even if we fail to model the experiments that would verify it by making too many concessions in the assumptions necessary for the experimental procedure to yield any results.
It may be seen as somewhat tautological in that if an experiment demonstrates it to be false, the economist will always reply that some of the assumptions of the experiment are untenable but, generally, I am inclined to believe that it would hold if the experiment were built in the correct fashion.
I somewhat agree with avm where she claims that exchange rates are determined by asset trading, which means that PPP would be a very poor predictor of exchange rates (and «that» is where the Big Mac index sheds some light, although it reflects prices on different goods, as I have mentioned earlier), but it should definitely constitute somewhat of a basic equilibrium condition on a hugely simplified model. The price index on tradables formulation should also prove to be correct if we can extract all possible impediments such as taxes, transaction costs, regulation and so forth that might distort the results.
In short, PPP is an important theoretical result that would definitely hold if we could expunge all imperfections from the actual prices of commodities. Not being able to do so, our only choice is a sensible interpretation of reality that does not violate a sound result in favour of experimental assumptions that are no less unrealistic.
Posted by: Pedro Serôdio at Feb 12, 2007 11:50:27 AM
is so ez indeed. eggsactly. its #8, all of the above (though none of the above would also be accepted for full credit). give the dining room set to Pete Boettke for me.
Posted by: kevin at Feb 12, 2007 12:24:04 PM
When I was last in London, I noticed that the prices seemed to be almost exactly what they'd be in the U.S.--except they were in pounds, which meant they were actually twice as high. Didn't price any dining sets, though.
Posted by: Virginia Postrel at Feb 12, 2007 2:18:11 PM
3, with a "depends what you mean" on 4 and then a couple "maybies"
9. Not a bad guide, at least in its relative version, to what happens in the long run
Posted by: younotsneaky at Feb 12, 2007 2:48:29 PM
re: DK, shouldn't any such "implicit tax" disappear into ground rent, at least in a simple economy with no transport costs?
Posted by: anon at Feb 12, 2007 2:52:13 PM
yes, anon, which is why i said "implicit tax or ...". If your simple economy doesn't include other regulatory costs, lawsuits, shortages of nearby workers due to rent control, etc., then yes it does disappear into ground rent.
and yes, josh, I'm with #1, tautology, but it seemed redundant to actually spell that out. #5 and maybe #4 are also appealing.
Posted by: DK at Feb 12, 2007 3:55:34 PM
(3) is correct.
(1) is false because PPP is falsifiable.
(2) and (4) are false because PPP does not always hold in the data,
and you do not define what is meand by good.
(5), (6), and (7) are false because although the real exchange rate is
highly volatile, it is (likely) mean reverting even when computed
from the CPIs (which includes differentiated products and
nontradables).
(8) is false because tautologies are not falsifiable.
Posted by: Martin at Feb 12, 2007 4:39:14 PM
(3) is correct.
(1) is false because PPP is falsifiable.
(2) and (4) are false because PPP does not always hold in the data,
and you do not define what is meand by good.
(5), (6), and (7) are false because although the real exchange rate is
highly volatile, it is (likely) mean reverting even when computed
from the CPIs (which includes differentiated products and
nontradables).
(8) is false because tautologies are not falsifiable.
Posted by: Martin at Feb 12, 2007 4:39:43 PM
As a theory, 2 and 3, as a hypothesis, 5 and 6.
However, it is falsifiable, and has found to be
false in situations where it would be expected to
be true, e.g. for identical goods right across
the US-Canadian border. This is a well-known
conundrum of international trade.
Posted by: Barkley Rosser at Feb 12, 2007 6:23:17 PM
3. Yes, since it does include implications for equilibrium exchange rates.
4. Yes, if that means that with the right simplifing assumptions in your equilibrium case you should get PPP.
5. Yes, since the theoretical explaination stems from the assumption of a single price for an individual homogenous good.
6. No, since it's only applicable to the index if applicable to the goods within the index. Indices help by averaging out short term variation, but don't have any theoretical significance.
Posted by: MattXIV at Feb 12, 2007 6:45:55 PM
Shouldn't correct answers get a big mac or I suppose lucky for us today, an iPod even?
Posted by: liberty at Feb 12, 2007 8:39:45 PM
2, 5 and 7.
Posted by: ricardo at Feb 12, 2007 9:41:42 PM
