Annals of “All research is wrong”

A popular study from the 1970s that helps sell millions of dollars’ worth of fish oil supplements worldwide is deeply flawed, according to a new study being published in the Canadian Journal of Cardiology.

The original study, by Danish physicians H.O. Bang and D.J. Dyerburg, claimed Inuit in Greenland had low rates of heart disease because of their diet, which is rich in fish oil and omega-3 fatty acids from eating fish and blubber from whales and seals.

“I reviewed this original paper and it turned out to be that they actually never measured the frequency of heart disease in [Inuit],” said Dr. George Fodor, the new study’s lead researcher.

Story and study.

Published medical science is deeply flawed. More often than not, when I’ve looked up a study claiming X, the statistics are deeply problematic. I suspect poor training and poor refereeing are proximately to blame, but there must be some deeper absence of incentives. It’s a shameful state of affairs.

Political science and economics are (a little) better, but as I teach my students, the first thing you should say to yourself as you open every book or research paper is, “This is almost certainly wrong.” Depressing but important. Welcome to science.

The long, LONG term impact of cash transfers to the poor

Nearly a century later…

we collected individual-level administrative records of applicants to the Mothers’ Pension program—the first government-sponsored welfare program in the US (1911-1935) —and matched them to census, WWII and death records.

Male children of accepted applicants lived one year longer than those of rejected mothers. Male children of accepted mothers received one-third more years of schooling, were less likely to be underweight, and had higher income in adulthood than children of rejected mothers.

A new NBER paper by Aizer, Eli, Ferrie, and Lleras-Muney. I do not see an ungated version on author websites but it may appear.

“Outbox vs. USPS: How the Post Office Killed Digital Mail”

An article in Inside Sources:

They wanted to allow consumers to digitize all of their postal mail so that individuals could get rid of junk mail, keep important things organized and never have to go out to their mailbox again. They set out to “redefine a long cherished but broken medium of communication: postal mail.” Customers would opt-in for $5 a month with “Outbox” to have their mail redirected, opened, scanned and available online or through a phone app.  Consumers could then click on a particular scanned letter and ask that it be physically delivered, or that certain types of letters not be opened (e.g., bills etc.).

…When Evan and Will got called in to meet with the Postmaster General they were joined by the USPS’s General Counsel and Chief of Digital Strategy. But instead, Evan recounts that US Postmaster General Patrick Donahoe “looked at us” and said “we have a misunderstanding. ‘You disrupt my service and we will never work with you.’” Further, “‘You mentioned making the service better for our customers; but the American citizens aren’t our customers—about 400 junk mailers are our customers.  Your service hurts our ability to serve those customers.”’

Hat tip to @owenbarder.

 

Princes and merchants

I love international development, but if I could do it all over again I might return to my roots as a political/economic historian. In the meantime, David Stasavage (among others) lives out my fantasy research agenda in a new paper:

…why did the autonomous city eventually die out as a form of state organization in Europe? The conventional explanation is that autonomous cities were economic powerhouses, but they died out because they could not compete militarily against larger states. It is suggested that this was particularly the case after technological change led to high fixed costs in war fighting. A problem with this argument is that autonomous cities long held a financial advantage over larger territorial states when it came to fighting wars. The autonomous cities found it easier to gain access to credit and at lower rates of interest, a feature that undoubtedly helped aid in their survival.9  My findings in this article point to a more simple reason why autonomous cities may have died out; the political institutions that initially fostered growth ultimately led to economic stagnation.

Can employment reduce lawlessness and rebellion?

In short, yes. A new paper by me and Jeannie Annan:

We evaluate a program of agricultural training and inputs to high-risk Liberian men, mainly ex-fighters engaged in illegal mining, logging, and rubber tapping. We show that increasing farm productivity raised the opportunity cost of crime and rebellion and led to more peaceful occupational choices. After 14 months, treated men shifted hours of illicit resource extraction to agriculture by 20%. When conflict erupted in neighboring Côte d’Ivoire, they were also less likely to engage in mercenary recruitment. Expected future payouts appear to be a further deterrent to mercenaries. We see no evidence the program affected occupational choice through peers or preferences.

The insight is not rocket science, but tracking these guys over two years was about as hard as putting someone on the moon. Thank God no one does the cost-benefit analysis of my research projects.

Here’s why I think the result matters. The theory makes sense–crime is an occupational choice the wealthier are less likely to make–but to date there hasn’t been any rigorous, individual-level evidence this is true. Armed violence is rare, and studies that capture it are usually observational and use country- or district-level data. Field experiments seldom study men with opportunities for crime or conflict. Unless you happen to be a naive glutton for punishment.

Also, from a policymaker’s point of view, it’s not clear that high risk young men who used to wield weapons want to beat their swords into ploughshares. The worry is that farming could deter rebellion in Liberia as well as MacDonald’s jobs stops drug dealing in Detroit. I suspected it to work, but I’m sincerely surprised just how well this program reduced illicit work and violence.

Corruption, revisted

One of the topics I’m most notorious for blogging about is the topic I know the least about: corruption. Basically, I’ve said it’s a second order issue, and a Western fetish. Here is a full slate of posts.

Matthew Stephenson, Harvard Law professor, has a new anti-corruption blog

As for citizens (especially poor citizens) in developing countries, if we want to know what they think about corruption, we could ask them–as indeed we have (see, for example, herehereherehereherehere, and here). And they consistently rank corruption as among their most significant concerns.

…I’ll throw out my own hypothesis about why so many academics in the blogosphere are drawn to the anticorruption-is-a-Western-obsession-that-doesn’t-matter-much-for-development canard: academics (and I speak as a member of the tribe) enjoy feeling like iconoclasts willing to speak uncomfortable truths to power. And in the development field, a certain type of academic particularly enjoys attacking anything that the major institutions (World Bank, U.S. government, OECD, etc.) seem to be for.

There are fair points throughout. I’ll make two in response.

  1. Second order issues are still important issues. The problem with most development policy is that everything is important. Everything matters for development. Setting priorities is hard. I haven’t seen the evidence that says corruption is one of the top 3 things the World Bank President or head of USAID should care about. Citizen surveys are useful, but not fully persuasive.
  2. Corruption strikes me as a symptom of deeper governance problem. Many if not most anti-corruption policies strike me as treating symptoms not the disease. Acemoglu and Robinson put it well: “corruption is a way for many economists and policymakers to talk about bad political outcomes without talking about politics.” I think we should be thinking about what builds bureaucratic capacity, how to restrain over-centralized regimes, and the political incentives leaders have to use patronage and repression to control their societies. Transparency might be part of that. And less corruption would be an important outcome. maybe. It’s not clear to me that either ought to be front and center policy aims.

Internship opportunity in Ethiopia

I’m looking for an intern to work on my “More Sweatshops for Africa?” project in Ethiopia, based in Addis. The position would be for ~12 weeks starting as soon as possible (ideally mid to late May). There is no salary but we cover travel, accommodation and living expenses.

This would be a great opportunity to get experience of how surveys are run in the field and to work in a fascinating country, Ethiopia. The job will be largely to help run three small but intense surveys and all associated administration. The intern will report to my in-country project coordinator.

Ideally the candidate would be a graduating or graduated college student with an interest in development, and hopefully some background in quantitative analysis (e.g. experience using Stata and MS Excel). MA students are welcome to apply but managing survey teams might not be using your skills effectively.

Resumes (1-page) and short statements of interest (1 page) should be sent here. Do not email me directly, since I won’t be handling the hiring process. My team will begin reviewing applications by Sunday 4th May at the latest, though later submission will be considered. Applicants will be contacted by mid-May if we are interested in speaking.

Deeply chilling sentences

I am starkly amazed: A very serious economics book on inequality, Thomas Piketty’s Capital, is currently Amazon’s #1 best selling book. Whether you like this book or not, the fact that it beats out Heaven is for Real should give you some hope for the country.

Before we get too giddy, a new paper by Derek Neal:

On any given day in 2010, almost one in ten black men ages 20-39 were institutionalized [in a prison or mental hospital], and rates of institutionalization were actually slightly higher among black men in 2000.

…[R]ates among less-educated black men have reached levels that were unthinkable prior to 1980. …[A]mong black men in 2010, more than 1 in 6 high school dropouts in their forties were institutionalized on any given day.

These numbers do not simply reflect a significant number of men who are serving extremely long sentences. Admission rates for new court commitments are now also noteworthy for this age group.

Punchline for the inequality averse: While we spare some time for Piketty’s global wealth tax, we could also think about ways not to imprison the majority of American black men in their lifetimes.

Today is a big day in the history of democracy

India is voting. To put things in perspective:

  • The number of Indians eligible to vote in the current election: ~815 million
  • The total voting age population in the US, European Union, Canada, Australia, and New Zealand: ~700 million

Other interesting factoids:

  • Putting the representative back in representative democracy, there are 543 constituencies, meaning each member of parliament represents roughly 1.5 million people.
  • Today is a big voting day, but the logistical challenges are large enough that voting is taking pace in nine phases between April 7 and 12 May.
  • Parties are expected to spend ~$500 million in the election, the world’s second highest after the US$7 billion spent on the 2012 U.S. presidential election.
  • The ruling Congress party is expected to have a historic and humiliating defeat, putting in power the Hindu Nationalist party.

I welcome other points and items.

For the India versus West voter numbers I thank Tavneet Suri and Roberto Rigobon. Their source.

Cash transfers: The index fund of international development

[This] illustrates another upside to cash transfers: they can serve as the index funds of international development. An index fund is a bundle of investments that is not actively managed, reducing the costs for investors. Its value simply reflects the upward and downward swings of the individual stocks that are included in the bundle. Similarly, a cash transfer is a development project stripped of any active management costs, and its performance tracks the success or failure of the individual recipient. Cash transfers thus provide a baseline for evaluating the active management performance of government officials and development professionals.

Move over Vanguard.

My article with Paul Niehaus, in the latest edition of Foreign Affairs (ungated if you register for free)

The “Amazon Tax”

From Brian Baugh, Itzhak Ben-David, and Hoonsuk Park:

Several states have recently implemented laws requiring the collection of sales tax on online purchases.  In practice, however, only Amazon.com has been affected.  We find that households living in these states reduce Amazon expenditures by 9.5%, implying an elasticity of -1.3.  We find the effect to be more pronounced for large purchases, for which we estimate an elasticity of -3.2. Further, we find that the decline in Amazon purchases is offset by a 2.0% increase in purchases at local brick-and-mortar retailers and a 19.8% increase in purchases at the online operations of competing retailers.

“Blow up the Times Op-Ed page, and start again”

From Alex Pareene at Salon, the tragedy of the New York Times is its terrible opinion section.

Don’t get me wrong, it is not as bad at the Washington Post’s opinion section, but the Times — specifically its slate of regular columnists — is bad. And that badness matters. The president and most of the sane members of Congress care what the Times and its opinion-makers say. The opinions of these columnists have consequences. But most of them don’t write like people who ought to be entrusted with that responsibility.

To which I can only say +1.

The CIA guide to sabotage

From the CIA archives:

This classified booklet described ways to sabotage the US’ World War II enemies. Many of the sabotage instructions guide ordinary citizens, who may not have agree with their country’s wartime policies towards the US, to destabilize their governments by taking disruptive actions…

  1. Managers and Supervisors: To lower morale and production, be pleasant to inefficient workers; give them undeserved promotions. Discriminate against efficient workers; complain unjustly about their work.

  2. Employees: Work slowly. Think of ways to increase the number of movements needed to do your job: use a light hammer instead of a heavy one; try to make a small wrench do instead of a big one.

  3. Organizations and Conferences: When possible, refer all matters to committees, for “further study and consideration.” Attempt to make the committees as large and bureaucratic as possible. Hold conferences when there is more critical work to be done.

Hat tip to SN.

This sounds like the United Nations. Coincidence?

Read the original manual.

The first control trial?

Apparently the first reference to comparing treatment and control groups is in the Book of Daniel in the bible, where King Nebuchadnezzar orders his people to eat only meat and drink only wine (for their health). When some object, he allows them to eat only legumes and drink only water for 10 days, after which he compares their health to the “control” group of meat-eaters. They do better, so he allows them to continue their diet. Note, however, the lack of randomization and hence the possibility of a strong selection effect. The king appears to implicitly understand this, since he does not order everyone to switch to the new diet.

 Via Julian Jamison.

Bleg: In the event you live in Geneva or Brussels or Nairobi

Since our daycare closes for 6 weeks this summer starting (don’t get me started) we are thinking of relocating to Brussels, Geneva, or Nairobi. Jeannie’s organization (IRC) has headquarters there, and I can plug away on papers from anywhere. And (in two of the three) work on my miserable French.

Here’s where the bleg comes in. We’re thinking July 28 to Aug 29 or even Sept 5. I figured there’s a small chance:

  • You live in one of these cities or know people that do
  • You or they might like to swap places. We have a lovely 2-bedroom (3 if you count a den) in Manhattan that can sleep 4-6.
  • Failing that, you have a place to rent a blogger and his lovable brood
  • In any of these cases, you know the perfect babysitter or nanny or daycare in either city (crucial to me plugging away on those papers)

Any other advice or thoughts on the cities welcome.

Unfortunately circumstances tie us to these three cities and so no substitutes. Though if you want to rent an apartment for August in New York, let me know.

I can be emailed at the usual spot.

Why am I an anti-Chapterite?

The most common comment on yesterday’s advice post for assistant professors: “I agree with everything except the tirade against book chapters.”

Let me say why I’m surprised, and ask for clarification.

  • Many people said “book chapters benefit me”. This is true. So does anything we do in our workday. The right question is “as opposed to what?” There is an opportunity cost to a book chapter. If you feel your scarcest resource is time, then shouldn’t you spend time on the projects that have the most net impact on your career and discipline?
  • If the answer is “I do! The marginal cost of a book chapter is lower than sending the same article to a peer reviewed journal, and the marginal benefit is about the same,” then (a) fair enough, but (b) I’m surprised, and (c) I suspect that times are changing. One reason being…
  • Unless you’re willing to buck copyright and post the chapter on your website (few people do) then a book chapter banishes most people from ever reading what you write–even other scholars. This is 2014. If your work isn’t accessible through a click or two on Google, it’s passed by more often than not.
  • Ergo, even if tenure practices in your department means that the private marginal benefit of a book chapter is high, the public one is low.

In sum: I fail to understand how edited volumes are a 21st century venue for knowledge.

On reflection, perhaps the solution is to make edited volumes more like online journals. Once again, however, we bash up against the fact that the academic publishers are still wallowing in 1965.

As an aside, it’s amazing to me that in just a year or two most academic blog conversations have shifted to Twitter. It’s more amazing to me that reasonable discourse actually happens in 140 characters. Of course, anything that stops academics from using 500 words where 30 will do might be a good thing.

Advice for new Assistant Professors

It’s a bit early for me to be giving advice (I’m only on the cusp of non-Assistant status), but I found myself asked for advice the other night by a table full of newly minted PhDs on to their first academic jobs. I also give advice, unasked, to my graduating students.

So, early or not, here’s what I passed on. Mainly of relevance, I suspect, to economists and political scientists.

  1. Learn to say no to new projects. Opportunities will start crossing your desk faster than you expect. It’s tempting to take the first ones, even though they’re likely the worst. There’s a big opportunity cost here: every project you take on now crowds out a potentially better one in a year or two.
  2. Have a higher bar for projects with big exit costs. It’s one thing to start a historical data collection project or a new theoretical model. You can always stick it in a file drawer if it goes poorly. But if you commit to a field experiment or a project with an eminent person, you are stuck with it to the bitter end. Make sure they are worth it.
  3. Book chapters and reviews are a waste of time. David Romer told us this in my first macroeconomics class, and I have come to agree. Few people read these, especially when they are buried in a $200 book no library buys. Unless you’re invited to do a Handbook chapter or an especially high profile book, it’s almost always better to put your article in a field journal. If it doesn’t merit publication in a decent field journal, probably it’s crowding out something more important to you and the world.
  4. Get your dissertation papers or book out. I see so many people too busy starting new projects to finish the old ones. This is the kiss of tenure death. Send menuscipts out soon after the job market, and make revisions your first priority when they come back.
  5. Seek out mentorship. Ask your dissertation committee and colleagues in your new department to read your abstracts and introduction, and strategize about framing, titles, and generally how to sell your work to a general audience. This is an art that takes years to learn, and personal advice can make a big difference in where you publish your early work.

A word of caution: almost all professional advice is either “here’s the mistakes I made” or “how to be more like me” in disguise. In this instance, it’s more of the former than latter.

For someone with real experience, you should read Greg Mankiw’s advice. I agree with all, except for “do not start a blog”, as “that will only establish your lack of seriousness as a scholar”. This may have been good advice in 2007 (when I defied the advice to start this website). Maybe it is good advice still and I don’t realize it yet. But here’s my estimate of the impacts so far:

  • Blogging has made me a better writer
  • It has meant I and my papers are much better known and cited by colleagues than otherwise
  • Opportunities cross my desk more often than otherwise
  • And, maybe most of all, I hold this blog almost directly responsible for several million dollars in research and program funding so far (paying for a lot of serious scholarly stuff). This number is exaggerated by the fact that I typically need to raise large sums for interventions as well as the research, but the basic point holds–for me blogging has (unexpectedly) paid back a hundredfold in scholarly work.

Yes there are costs and risks, but I think social media is too important for young academics to ignore. Accordingly:

  1. If you want to tweet or blog under your research name, be serious. Let your research interests influence your blogging. Become a professional resource for people in your subfield. Be constructive and thoughtful not critical, and never use social media to attack colleagues. This will be a public good that pays back privately.

Colleagues: please add your advice below.

“he flourished his sword, I pulled out my gun and shot him, and then we went back to England”

Harrison Ford does an “Ask me anything”

We were shooting in Tunisia, and the script had a scene in which I fight a swordsman, an expert swordsman, it was meant to be the ultimate duel between sword and whip. And I was suffering from dysentery, really, found it inconvenient to be out of my trailer for more than 10 minutes at a time.

We’d done a brief rehearsal of the scene the night before we were meant to shoot it, and both Steve and I realized it would take 2 or 3 days to shoot this. And it was the last thing we were meant to shoot in Tunisia before we left to shoot in England. And the scene before this in the film included a whip fight against 5 bad guys that were trying to kidnap Marian, so I thought it was a bit redundant.

I was puzzling how to get out of this 3 days of shooting, so when I got to set I proposed to Steven that we just shoot the son a bitch and Steve said “I was thinking that as well.”

So he drew his sword, the poor guy was a wonderful British stuntman who had practiced his sword skills for months in order to do this job, and was quite surprised by the idea that we would dispatch him in 5 minutes. But he flourished his sword, I pulled out my gun and shot him, and then we went back to England.