« "Bush Midsession Budget: Profound Sadness" | Main | Arnold Kling Doesn't Understand »

Tuesday, July 29, 2008

"No Time to Think, Liberalize!"

Not too long ago, I posted a Vox EU article on the controversy over how large the benefits from trade are for the U.S. Josh Bivens of the EPI is mentioned in the article, and he would like to respond:

No time to think, liberalize!, by Josh Bivens: Hufbauer and Adler's VoxEU piece doesn't really advance the ball much further down the field in this debate. But, just to sum up for any interested readers: a paper by Bradford, Hufbauer, and Grieco (BGH, henceforth) was released in 2005 that had estimates of the US gains from trade liberalization that were far, far outside those estimated by previous research: they claimed past trade liberalizations had added almost a trillion dollars to the US economy by 2004, and, future liberalizations offered the promise of adding another half-trillion.

Dani Rodrik didn't think much of it. And, after seeing these numbers used to great effect in the political debate, I wrote a long-ish working paper and two associated issue briefs detailing why they were unreliable. My over-arching critique was that their study was a literature review that cherry-picked the research for the absolute maximum gains that could be attributed to liberalization while ignoring any reasons (even within this same literature) as to why these gains might be much, much smaller. In the end, I found nothing to shake my belief (based on staid, mainstream economics) that the real gains from liberalizations are closer to a tenth or less of what they claim.

Given that I was named in the Hufbauer/Adler piece, I feel obliged to respond, but, am hard-pressed to find much to say given that their piece mostly just describes the approach taken in the original paper (which I still don't like, but, the papers linked above are where readers should go if they care why), and, their specific pushback against my own work consists mostly of snark and non sequiturs.

On the first, it's a little unclear why the back-of-the-envelope calculation of the gains from trade liberalization that I used is "simple-minded". It's textbook economics, after all, and, it yields an estimate of the gains from trade that is quite close to that provided by the World Bank study cited by Rodrik, and, it even provides a pretty liberal estimate compared to what is identified by the (behemoth) USITC study on the Economic Effects of Significant Import Restraints. I guess they just like the insult.

As to the non sequiturs, they argue that I "ignored Solow's landmark finding (1956) that productivity gains explain above 80% of US economic growth". They're right, I did. I am familiar with this finding; I know, for example, that it was actually from Solow (1957), not Solow (1956), and, much more importantly I know that it is utterly irrelevant to the debate at hand. If there's a convincing link between the arguments of BGH and Solow, it sure wasn't made in the original paper, which includes no citation of Solow in its 5-page bibliography. This argument gives hand-waving a bad name.

Hufbauer and Adler also complain that my critiques in the longer paper are "repetitive". Well, yes, because it was reviewing the BGH paper which made the same mistake over and over again: cherry-picking the studies they review for the maximum number they can pin on the gains from trade liberalization while ignoring anything that cuts in the other direction.

The VoxEU piece does offer one significant change in emphasis relative to the BGH work - Hufbauer and Adler argue that "Rapidly falling transportation and communication costs are perhaps more important features of the globalisation story [than policy liberalization]". The original BGH piece claimed that all future gains it identified stemmed solely from policy liberalization. It was precisely claims about the scope of gains from policy liberalization that were so hard to swallow, especially those regarding future payoffs we should expect from signing ever more trade agreements.

To give one example why, table 2.4 in the original BGH paper gives the most transparent accounting of their claims regarding a major "policy-only" impact of trade, and, even given their very generous interpretation of the study upon which the calculation is based, what it shows is that between 1982 and 2002 trade liberalization added all of $1 billion to the US economy.

This is twenty years that saw the signing of the North American Free Trade Agreement (NAFTA), the completion of the Uruguay Round of the General Agreement on Tariffs and Trade (GATT) which led to the formation of the World Trade Organization (WTO), and the permanent normalization of trade relations with China as well as its entry into the WTO. $1 billion seems like a pretty small payoff for twenty years of liberalizations that were politically wrenching, and, it seems like an awfully thin reed upon which to claim that future liberalizations will somehow yield payoffs exponentially larger.

Hufbauer and Adler (and other proponents of continuing the globalization status quo) often hold out the service sector as an explanation for how future liberalizations will yield much greater benefits than past ones. Anything's possible, I guess, but, service-sector trade as a share of US GDP is still less than a third as large as goods trade, and, more importantly, it is awfully hard to measure just how protected the service sector is.

The studies they cite regarding future benefits from service sector liberalization generally take the full value of price (actually gross margin) differences between service sector output in the US and trading partners as de facto evidence of barriers to trade in services. This just won't do - there are plenty of reasons for prices to diverge that have nothing to do with protectionism, and, assigning the full value of the difference to trade protection is, again, just cherry-picking for the highest number one can get regarding the prospects for trade liberalization to improve US incomes.

Probably needless to say at this point, I also disagree that empirical research shows no (or a trivial) role for trade flows in influencing income distribution. The more immediate point, however, is that we have time to get this debate and our response to globalization right: threats of huge income losses if we don't all get on board the next trade agreement that comes down the pike are empty, and the golden goose won't wander away if, say, the Colombia/US agreement or any other agreement is allowed to wither and die on the vine.

    Posted by on Tuesday, July 29, 2008 at 01:17 PM in Economics, International Trade | Permalink  TrackBack (0)  Comments (23)


    TrackBack URL for this entry:

    Listed below are links to weblogs that reference "No Time to Think, Liberalize!":


    Feed You can follow this conversation by subscribing to the comment feed for this post.