You will find it here. The contents include:
James Tooley on Abhijit Banerjee and Esther Duflo’s Poor Economics: Banerjee and Duflo propose to bypass the “big questions” of economic development and focus instead on “small steps” to improvement. But, says Tooley, they proceed to make big judgments about education in developing countries, judgments not supported by their own evidence.
Why the Denial? Pauline Dixon asks why writers at UNESCO, Oxfam, and elsewhere have denied or discounted the success and potentiality of private schooling in developing countries.
Neither necessary nor sufficient, but… Thomas Mayer critically appraises Stephen Ziliak and Deirdre McCloskey’s influential writings, particularly The Cult of Statistical Significance. McCloskey and Ziliak reply.
Was Occupational Licensing Good for Minorities? Daniel Klein, Benjamin Powell, and Evgeny Vorotnikov take issue with a JLE article by Marc Law and Mindy Marks. Law and Marks reply.
Mankiw vs. DeLong and Krugman on the CEA’s Real GDP Forecasts in Early 2009: David Cushman shows how a careful econometrician might have adjudicated the debate among these leading economists over the likelihood of a macroeconomic rebound.
Kinda sad that Cushman writes an entire paper without ever becoming aware of the literature on Okun’s Law…
Are you sure these are people you want to direct your readers to?
Yours,
Brad DeLong
P.S.: Tooley seemed to me to be substantially off-base…
So, what is the real reason you were wrong about the magnitude of the rebound (or do you deny that you made a mistake)?
P.S.: Tooley seemed to me to be substantially off-base…
That’s hardly helpful unless you deign to enlighten us how. I read the article and Tooley has many valid points.
His statement is typical of the usual defense made by ideologues. Logic usually takes a back seat, while opinions – free from the encumbrance of reason – are usually expressed freely and forcefully.
This is especially ironic because this pattern aligns with Tooley’s own criticism of the book he reviews – that its “big picture” proposals are not supported by the evidence discussed. DeLong is just providing further evidence that in support of Tooley’s thesis.
The DeLong/Krugman support of the CEA forecast was based on two simple assumptions: (1) the high level of unemployment in 2009 will rebound to its historical average by 2013 and (2) Okun’s Law holds. Which of these failed to hold? Clearly unemployment did not rebound. One paper by Cushman, which uses statistical techniques to limit the sample to 1986-2009, is not going to convince anyone that a robust empirical macro result like Okun’s Law no longer matters.
And when it comes to evaluating why unemployment hasn’t rebounded, of course we’re back to the same old debate: is unemployment is high because of structural employment, or because monetary/fiscal policy have been historically bad at stimulating aggregate demand?
The former. Demand existed before monetary and fiscal policy were invented. To wake it up again you must remove the structural obstacles.
I tweeted. John B Taylor testified. And the employment of federal workers engaged in regulatory activities absolutely soared.
And state and local government employment tanked…what’s your point? I don’t see the financial crisis/housing bubble bursting as a big victory for unfettered markets and there were legitimate concerns about systemic risk to address in the aftermath. Governments are not the only source of uncertainty (or other structural impediments) for households and firm. And I think it’s hard to argue that the productive capacity of the US workforce decline so precipitously in the fall of 2008. A rise federal regulators off a modest base isn’t a compelling argument to me.
Re Cushman, I think his paper is more limited than you think: he’s just trying to prove Mankiw was right and Krugman/DeLong were wrong on the rebound. I like this fn #4: “In the interest of fuller disclosure, I’ll reveal that more recently, in September 2011, I received (in contrast to pleasant Mankiw emails) an unpleasant email from DeLong. It was in response to a comment I had attempted to post on his blog. He wrote me, “Shame on you for trying to confuse the issue.” He also did not allow the comment to be posted. As my project was already well underway, the event was clearly not an initial motivator, but may have served to spur me on. In case the reader wants to assess whether I was “confusing the issue,” I provide details in the Appendix. The episode also provides a case study of the sorts of things I don’t like in DeLong’s and Krugman’s blogs. ” – No fury like a blog poster scorned!
A simple ‘oops, look like I flubbed that one.’ would suffice.
Yes, Brian because that’s how you sign off on threads you’ve lost.
Neither Mankiw nor DeLong nor Krugman signed off on the policy forecasts and quite frankly the personification of such forecasts is absurd and beside the point. The macro question is why did the economy do what it did…not who figured it out first. Making lucky guesses or being right for the wrong reasons does not get you any points…except it seems in the blogosphere.
‘A fall in production that is accompanied by a big rise in the unemployment rate will in all likelihood by reversed.’ – Brad DeLong, 2009
if I had expressed a (characteristically strong, confident Keynesian-style) utterance such as above, I would own such a statement now.
I would hope my reaction would not be haughty dismissal of the criticism.
Of course, i would not adopt the breezy 2009 DeLong style in the first place, for fear that pople might great the mistaken impression that I was practicing something akin to science.
Brad, you didn’t know, you don’t know, stop with the “in all likelihood” crap.
Counterpoint: Tooley seemed to me not to be substantially off-base.
I expect there will be many, many analyses like Cushman’s which try to unpack the forecasts of key policy makers in the recession / recovery. I don’t think we learned here that Mankiw was right because there’s a unit root in GDP. Personally, I think you need to get more ‘under the hood’ of these forecasts to see why they differed from each other and reality. As one example, fiscal and monetary policy assumptions could have differed both in their details and assumed effects. I do think this is a interesting question (though I wouldn’t have framed it around a blogger debate), but this paper just gives one bit of the answer.
“Tyler” has just been blamed for the fact that – - Mankiw vs. DeLong and Krugman – - is trending on twitter.
https://twitter.com/heathbarrr_/status/249688549174169601
Could be like the Facebook Dutch party which sparked a riot yesterday as reported worldwide.
“Was Occupational Licensing Good for Minorities?”
The discussion seems to be about whether occupational licensing was more detremental to minorities than to others. I did not see any suggestion of a calim that occupational licensing was good for anyone.
It is almost always the case that it is good for incumbent licensees, hence their support for it.
The discussion is more: was the ability to become an incumbent discriminatory across race and gender lines, at a statistically significant level.
I thank Brad DeLong for his characteristic brand of analysis in his comment above. Negative reactions can be useful, even if unpleasant. The literature on Okun’s Law shows it to be a rather unstable relationship (e.g., for something recent, see “How Useful is Okun’s Law,” by Edward Knotek, Kansas City Fed Economic Review (2007), http://www.kc.frb.org/publicat/econrev/PDF/4q07Knotek.pdf). My paper noted that Profs DeLong and Krugman relied on Okun’s Law, but my point was they did not do so in a way suitable for meaningful forecasting. My point was not the provide a discussion of Okun’s Law. In fact, I took their suggestion to include unemployment and not just GDP growth (giving me the two Okun’s Law variables) quite seriously. The VAR forecasting analysis, which is half the paper’s forecasting analysis, is, in fact, a dynamic Okun’s Law model. And anyone who reads the paper will also know I had to deal with the issue of stability, or lack thereof.
If you’re Brad De Long you’re never wrong.
Stop that Rich, you’ll never become a [b]burgher[/b] if you diss the literati.
Herr Professor-
Have mercy on me, a mere commenter!
Thanks for directing me to James Tooley critique of the PE chapter on education RCTs.
I see a similar pattern among randomista studies on interventions to improve access to and/or quality of health services and products in developing countries. The need or desire to answer “big questions” or, pretend to policy relevance (and levels of external validity they lack) means that RCT papers in health very frequently have conclusions that go (way) past the statistical findings. And to generate these conclusions, researchers mostly just pull up some scattered references to support their own biases – as Banerji and Duflo did in their education chapter.
Many many RCTs on health interventions measure demand curves. And, they find that, even for people in poor countries(!) and for stuff that is really good for them, DEMAND CURVES SLOPE DOWNWARD. One imagines having this as your key finding might make it hard to get your paper published. So, the authors leap from that to lay out conclusions with policy relevance – often commenting, as James notes on “big questions” which their research didn’t touch on. A common leap in health RCT papers is: our analysis confirms that demand curves slope downward…SO government should provide (whatever) for free through it’s facilities/ supply system.
The biggest insights from these papers are really about the authors pre-existing beliefs about how to get services or products delivered sustainably in developing countries.
I appreciate that James took the time to draw this out. I hope it helps your readers to be on the alert when reading RCT studies on how developing countries can improve their health or education outcomes.
Comments on this entry are closed.