Experience 1

Submission of
"No Razor’s Edge: Reexamining Alwyn Young’s Evidence for Increasing Inter-Provincial Trade Barriers in China."

Robert Barro came back with three referee reports --- the numbering scheme suggests the existence of four --- and all three of them come across to me as fair. Since they are all fair, I am reluctant to publish them here. (They are also all positive.)

Robert Barro in his accompanying e-mail writes:

Date: Wed, 07 Jun 2006 15:35:48 -0400
From: Robert Barro <rbarro_qje@gemini.econ.umd.edu>
Subject: QJE: decision on your submission
To: Carsten Andreas Holz <socholz@ust.hk>

Dear Mr./Prof. Carsten Holz:

Thank you for sending us this interesting paper, MS 11609, entitled
"No Razorâ?™s Edge: Reexamining Alwyn Young's Evidence for Increasing Inter-Provincial Trade Barriers in China"
for publication in the Quarterly Journal of Economics.

I have spent a lot of time thinking about Young's paper and your comment on his analysis.  I am convinced that Young has a lot of shortcomings.  However, I am particularly persuaded by one of the referees that Young's paper is now well-known to have problems; for example, the Bai et al paper in JIE does a better job on output convergence.  Also, as pointed out by another referee, you do a better job of criticizing Young than of coming up with sound alternatives.  In the end, I cannot justify the space in QJE to publish your critique, so I have to reject your paper.

By the way, I think your argument about price dispersion is incorrect in part; Young uses the LOG of P, so the coefficient of variation for this is not meaningful.

Sincerely yours,

Robert Barro
Editor
qje_admin@gemini.econ.umd.edu
The Quarterly Journal of Economics

To which I replied:

Date: Wed, 28 Jun 2006 04:46:32 +0800
From: Carsten Holz <socholz@ust.hk>
Subject: Re: MS 11609
To: Robert Barro <rbarro_qje@gemini.econ.umd.edu>

Dear Professor Barro,

I wonder if there is any scope for a reversal of your decision to reject my manuscript "No Razor's Edge: Reexamining Alwyn Young's Evidence for Increasing Inter-Provincial Trade Barriers in China."

In your e-mail (appended below) you provide two reasons for the rejection and I wish to argue that both of your arguments are not valid.

A. You write, first, that you are "particularly persuaded by one of the referees that Young's paper is now well-known to have problems; for example, the Bai et al paper in JIE does a better job on output convergence."

(1) Bai et al. cover one of AY's four quantitative points; they do not cover the other three quantitative points, nor his qualitative analysis. If this one point were AY's key argument, then 6 pages would have sufficed for AY's original QJE article, rather than 46. AY himself clearly does not think that this one particular quantitative analysis suffices: "I have shown... To establish the argument, however, I need to present evidence of a movement of factor allocations away from patterns of comparative advantage..." (p. 1119, leading to this fourth quantitative point). I deal with all of AY's five points, including with his all-comprehensive final quantitative one.

(2) R1 writes "I don't think that Young's paper is now taken very seriously by China scholars." This is one out of three referees, and I point out the word "China" in front of "scholars." What percentage of the QJE's readers are China scholars?

(3) China scholars so far had no compelling counterargument to AY's argument besides the criticism of the first and second quantitative analysis. Perhaps, as in my case, their gut feeling simply said this is not right. (One of the appendices that I posted on the webpage that I set up for this paper, at http://ihome.ust.hk/~socholz/NoRazorsEdge.html, tells an experience in China where much of my gut feeling comes from.) I would think that a formal, compelling counterargument is of different quality than gut feelings.

(4) I just read through Under New Ownership: Privatizing China's State-Owned Enterprises, Yusuf/Nabeshima/Perkins, published by the World Bank/Stanford University Press, 2006, to read on p. 11 "According to Alwyn Young (2000), partial integration is costly for China and detracts from full exploitation of regional comparative advantages." So AY's paper is taken seriously even by (at least some) China scholars. Browsing the Web of Science (SSCI), I find 3 citations of his paper in 2006 so far, all taking his article seriously.

(5) While I was on sabbatical leave to Stanford in spring 04 I sat in on a graduate-level macro/ growth course, where AY's paper was taken as seriously as any other paper. This signaled to me that non-China scholars are in urgent need of a formal counterargument, and it was the trigger for me to write this paper.

B. Your second reason is that I, "as pointed out by a referee, ... do a better job of criticizing Young than of coming up with sound alternatives." I assume this refers to R3's comment "The quantitative part of this paper by itself constitutes a good comment. The major weakness of this paper is that the alternative explanations the author offers to explain what we observed in the data are not very convincing and coherent." The referee on p. 2 notes that the paper tries to do 3 things and "does a fair job for 1, a good job for 2, while a less convincing job for 3 [the alternative explanations]."

I don't need the alternative explanations to make my point that none of AY's quantitative evidence of increasing local protectionism holds. (a) Across all four quantitative analyses, I show that the U.S., following AY's logic, had increasing inter-state protectionism over time (in the third analysis limited to some periods). (b) For the first, third, and fourth quantitative analyses, my alternative evidence suggests AY has it wrong; I could just drop the weaker ones of my alternative explanations. On the third quantitative analysis, where I feel my alternative argument is the strongest point against AY findings, R4 writes "Furthermore, Holz argue[s] convincingly that the observed cyclical patterns for monthly industrial price data match relevant institutional changes during the 1980s and 1990s, which may not be caused by local protectionism."

If I take Popper's view of how science progresses, through bold conjectures and falsification (or non-ability to falsify), AY's paper is a bold conjecture. I set out to falsify it. I think I have a rather compelling case. R4 says "Holz has done more than the above summary, but those are the most significant findings I found very compelling, more compelling than those of Young." I don't need to provide an alternative explanation of what is being observed, when I can show that what AY observes does not hold up in more careful analyses of the evidence (as I do for the first, third, and his final, overarching fourth argument). Only in the second quantitative analysis can I not falsify his findings outright but rely on institutional changes to explain the temporary price divergence (AY has no evidence here, of long-run increasing protectionism, to begin with).

You point out as an aside that my "argument about price dispersion is incorrect in part; Young uses the LOG of P, so the coefficient of variation for this is not meaningful."

I am not sure if I get your point. AY uses the log of the standard deviation of the log of P. (If the coefficient of variation in the log case is not meaningful, is the standard deviation, as used by AY, meaningful?) I do *not* use the coefficient of variation of the LOG of P. I use the coefficient of variation of P, plain and simple.

Looking through the three referee reports, I view them as rather positive, ranging from "very compelling" (R4) to "this leaves me somewhat on the fence about the paper" (R1). I feel able to work with the referee comments, either address them (most cases), or make an argument as to why not to address them. The only item I can definitely not address is R3's suggestion to develop a (presumably mathematical) intra-industry trade plus pricing-to-market model; I simply don't have the expertise for that.

Sincerely,
Carsten Holz

which yielded a:

Date: Fri, 14 Jul 2006 19:20:50 -0400
From: Robert Barro <rbarro@harvard.edu>
Subject: qje
To: socholz@ust.hk

Prof Holz:

I spent a lot of time with your paper for QJE and there is no further scope for
consideration.

Robert Barro

that concluded the matter.

Note: after spending "a lot of time" with my paper, the one comment Professor Barro offers himself (the log P issue) is, in my view, not a criticism of my paper, as intended, but a criticism of Alwyn Young's paper that was previously published in the QJE.
 

I also sent my paper to the NBER, which earlier published Alwyn Young's paper as a working paper. This is our correspondence:

Date: Tue, 27 Jun 2006 16:49:06 -0400 (EDT)
From: Carsten HOLZ <socholz@ust.hk>
To: webmaster@nber.org
Subject: NBER Webserver Comments

Hi,

the NBER working paper series contains a working paper by Alwyn Young on distortions and incremental reform in the People's Republic of China, NBER working paper series no. w7828.

I have examined all of Alwyn Young's five arguments in his NBER working paper, one by one, and found none to hold up to scrutiny. This was a lengthy project and resulted in an academic paper on its own, directly and fully addressing this particular work by Alwyn Young.

I wonder if there is any possibility for a non-NBER researcher to make my contrary voice heard to the same audience as Alwyn Young could with his original paper. I.e., is there any possibility to have my paper published as an NBER working paper?

The paper is posted at http://ihome.ust.hk/~socholz/NoRazorsEdge.html

Sincerely,
Carsten Holz
(Associate Professor, Social Science Division, Hong Kong University of Science & Technology)

with the following response

Date: Wed, 28 Jun 2006 09:45:51 -0400
From: Gerri Johnson <gjohnson@nber.org>
Subject: Fwd: NBER Webserver Comments
To: socholz@ust.hk

Thank you for your interest in the National Bureau of Economic Research Working Paper series.  Our rules are that Working Papers are limited to NBER Research Associates and Faculty Research Fellows and to papers prepared for NBER conferences which are to be published.
 
 

Experience 2

Submission of
"China's Economic Growth 1978-2025: What We Know Today about China's Economic Growth Tomorrow."
(as of 26 Dec. 06, the above version is outdated; see my homepage http://ihome.ust.hk/~socholz for the most recent version)

Robert Barro responds by letter, 14 July 2005:

I am sorry to report that we will not be able to publish your paper, "China's Econmic Growth 1978-2025: What We Know Today about China's Economic Growth Tomorrow." I have concluded that there is not enough value added in the present paper to warrant publication in the QJE, and am returning a copy to you.

As it so happens, the QJE vol. 111, no. 6 (Dec. 2003) published an article by Alwyn Young entitled "Gold into Base Metals: Productivity Growth in the People's Republic of China during the Reform Period." This article discusses Chinese output, capital, and labor/education data and comes up with a TFP estimate.

My growth paper deals with the same variables and similar calculations --- except that I think it does it better --- and it does most of it in two lenghty appendices because I didn't think there is enough value added in such calculations to warrant publication in the QJE. (One of the appendices, on capital, after adding significantly more value, turned into a separate article.)

In my growth paper, instead, I show briefly why a TFP estimate for China (such as Alwyn Young calculated as the climax of his paper) can only be obtained by making assumptions that are not supported by the data (the assumptions are implicitly made by Alwyn Young), and otherwise get at the growth topic through a number of alternative channels. -- I have now taken the production function estimations [TFP issue] out of the paper because a referee at another journal, out of 50 pages, saw only two and was offended by the findings.