Chad Syverson, “Market Structure and Productivity: A Concrete Example”, the Journal of Political Economy (2004).

Notes: There’s a burgeoning interest in development in the question of why large productivity differences can persist across firms. One story is that they persist because of capital market imperfections—-the inability of capital to flow towards the more productive firms and away from the low productivity firms, putting them out of business. Syverson shows, using data on the ready-mix concrete industry in the US, that productivity differences can easily arise from demand side forces as well. His title also makes a really first rate pun.

Abstract: Many studies have documented large and persistent productivity differences across producers, even within narrowly defined industries. This paper both extends and departs from the past literature, which focused on technological explanations for these differences, by proposing that demand-side features also play a role in creating the ob- served productivity variation. The specific mechanism investigated here is the effect of spatial substitutability in the product market. When producers are densely clustered in a market, it is easier for consumers to switch between suppliers (making the market in a certain sense more competitive). Relatively inefficient producers find it more difficult to operate profitably as a result. Increases in substitutability truncate the productivity distribution from below, resulting in higher minimum and average productivity levels as well as less productivity dispersion. The paper presents a model that makes this process ex- plicit and empirically tests it using data from U.S. ready-mixed concrete plants, taking advantage of geographic variation in substitutability created by the industry’s high transport costs. The results support the model’s predictions and appear robust. Markets with high demand density for ready-mixed concrete—and thus high concrete plant densities—have higher lower-bound and average productivity levels and exhibit less productivity dispersion among their producers.

James J. Heckman, “The American Family in Black and White: A Post-Racial Strategy for Improving Skills to Promote Equality”, NBER Working Paper (2011)

Notes: This new paper is an excellent overview of some stylized facts about inequality  from a Nobelist who has spent years studying the topic. There is substantial inequality. Heckman reports that black males (females) make 25% (17%) less annually than whites males, Hispanics males (females) 15% (7%) less. He argues, however, that this cannot be due to discrimination by firms. Controlling for early childhood test scores, these gaps almost disappear. Indeed, black and Hispanic females actually make substantially more than whites once controlling for these test scores. Similar patterns can be found for other outcomes we care about, such as health, schooling attainment and incarceration status. Once you control for early child skill levels, inequality in these outcomes, for the most part, goes away. Strikingly, controlling for child test scores, minorities often do better; blacks and Hispanics become 15 percentage points more likely to go to college.

Heckman concludes with an argument for policy to support parents and early child development, which you’ll find in the following abstract. The literature here on which policies actually work is less developed, but Heckman makes a strong case that at least the type of policy required is clear.

Abstract: “In contemporary America, racial gaps in achievement are primarily due to gaps in skills. Skill gaps emerge early before children enter school. Families are major producers of those skills. Inequality in performance in school is strongly linked to inequality in family environments. Schools do little to reduce or enlarge the gaps in skills that are present when children enter school. Parenting matters, and the true measure of child advantage and disadvantage is the quality of parenting received. A growing fraction of American children across all race and ethnic groups is being raised in dysfunctional families. Investment in the early lives of children in disadvantaged families will help close achievement gaps. America currently relies too much on schools and adolescent remediation strategies to solve problems that start in the preschool years. Prevention is likely to be more cost-effective than remediation. Voluntary, culturally sensitive support for parenting is a politically and economically palatable strategy that addresses problems common to all racial and ethnic groups.”

(Source: papers.nber.org)

Philippe Aghion, Christopher Harris, Peter Howitt and John Vickers, “Competition, Imitation and Growth with Step-by-Step Innovation”, Review of Economic Studies (2001).

Philippe Aghion, Nick Bloom, Richard Blundell, Rachel Griffith and Peter Howitt, “An Inverted-U Relationship”, Quarterly Journal of Economics (2005).

Notes: Competition is frequently useful. In many markets, competition between producers keeps consumer prices down. Competitive bidding for government contracts reduces the cost of projects to the taxpayer. Whether competition is good for growth, however, is the subject of great debate. Let’s take as given that growth is driven, in some part, by innovation—-firms inventing new products. We have good reason to think competition is harmful here. Indeed, we use the patent system to allow inventors to gain monopoly profits as the sole vendors of their inventions. This may hurt consumers, but it gives them incentives to invent new products in the first place.

In these two papers, the authors show that this issue is more nuanced than it seems. They argue that what gives inventors incentives to innovate is not the absolute level of profits from inventing, but rather the benefits of inventing relative to being regular business people. If an industry is not competitive, being a regular business person is not so bad. As Hicks wrote, “the best of all monopoly profits is a quiet life”: You can charge higher prices to your customers and profit nicely. Why bother inventing if life is already so good? Microsoft is a great example, perhaps; has anything since Windows 95 really been very useful? On the other hand, if your industry is competitive and your margins are thin, the benefit of innovating and escaping competition by being the monopoly seller of a new product is quite high.

In UK data the authors find a U-shaped relationship, consistent with their theory. In industry years with low levels of competition, there are few patents, consistent with the “quiet life” hypothesis. In industries with high competition, there are similarly few patents, consistent with our first story. In industries with middling competition, there are the most patents. Their story argues for moderation: If you want growth in an industry, have competition. But, if it is easy for people to copy each other’s ideas, you want a little less; if it is difficult to do so, you want a little more.

Emerging Capital Partners, based in DC, is the largest Africa-focused private equity fund, with over 1.8 billion dollars invested in the continent. I listened to their co-CEO, Hurley Doddy, speak on a panel at the Harvard Business School’s Africa Business Conference in February. ECP makes a unique argument for investing in Africa. Not only do they argue you can get fast growth and above market returns (they claim between 20-30%), but they argue you get investments that are uncorrelated with returns in the US and Europe. High return, uncorrelated assets (low beta in the language of CAPM), of course, are exactly what investors are looking for to diversify their portfolios.

I did a quick review of the growth literature and there doesn’t seem to be a paper looking at this question. How correlated is growth in Africa with growth in the rich economies? How correlated is it with growth in other economies? The recent flight of capital from emerging markets was predicated on a belief that returns in the Middle East are correlated with returns elsewhere. It would be a worthwhile exercise to find out to what extent this is true.

Dean Karlan and Jonathan Zinman, “Observing Unobservables: Identifying Information Asymmetries with a Consumer Credit Field Experiment,” Econometrica (2009)

Joseph Stiglitz and Andrew Weiss “Credit rationing in markets with imperfect information,” American Economic Review (1981)

Pierre‐Andre Chiappori and Bernard Salanie, “Testing for Asymmetric Information in Insurance Markets,” Journal of Political Economy (2000)

Notes: When two people want to sign a contract, but one has more information about their circumstances than the other, economic theory predicts they’ll have two problems: adverse selection and moral hazard. Credit markets provide a classic example. Suppose borrowers know more than lenders about their ability to pay back their loans (e.g. the yield they’ll get from their farm after using the fertilizer they’re planning to buy on credit). The adverse selection problem was first shown by Stiglitz and Weiss (1981). The lender has the problem that the higher rate the rate she offers, the more high risk people (“low types”) will be attracted to her pool. This is intuitive. If you, the borrower, know you’re likely not to pay anyway, you’re willing to take a higher rate. If you’re likely to pay, you want a lower one. Since the lender fears pulling these people into her pool, she restricts the amount of loans she gives out. Ultimately, some “high types” that would have paid back will not have access to credit. The other problem is moral hazard. Once the borrower gets the loan, she knows she won’t reap all the rewards (e.g. 50 percent of the harvest goes to the lender, in interest) and so she works less hard. The probability of default goes up. If the lender anticipates this, she should restrict the supply of credit, and provide less credit than in the situation in which she could make sure that the borrower gave the work his all.

One imagines that these problems are both important in poor countries, where we think that many people with high return investments might not have access to the credit they need to realize them. The tough bit is that they’re hard to study. How do we know which is going on, when they both yield the same result?

Karlan and Zinman (2009) do an interesting experiment in South Africa to see which of these problems is more important. They convinced a bank to send out direct mailers to people, offering them loans at randomly selected interest rates. Once certain people responded, having read the rates in the mailer, the bank yet again randomly selected rates and proposed these new rates to the potential borrowers. Comparing people who selected into loans on the first rate they were offered with those actually took the loans (and presumably invested in trying to pay back or not) at different rates, the authors can separate out moral selection and adverse selection. This is a really smart paper, and one of the few very rigorous advances in empirical tests of contract theory since Chiappori’s and Salanie’s (2000) investigation of French auto insurance markets. These market failures are difficult things to test for precisely because they rest on information that is not observable to researchers.

What is perhaps most interesting about this paper though, is that the effects of both adverse selection and moral hazard are so small. Adverse selection appears to have no effect on default rates, and the effect of moral hazard is tiny, around 2 percentage points. Chiappori and Salanie similarly cannot reject the hypothesis that there is no moral hazard or adverse selection in their insurance market. This makes us wonder, then, whether these problems are really first order problems after all.

Michael Kremer, Jessica Leino, Edward Miguel, Alix Peterson Zwane “Spring Cleaning: Rural Water Impacts, Valuation and Property Rights Institutions,” forthcoming in the Quarterly Journal of Economics (2011)

Notes: This paper from my current professor (Kremer) and former boss (Miguel) and co-authors gives us an empirical case for the value of communal property rights institutions over private property rights institutions in a certain setting (water springs in Kenya). The paper also demonstrates two important empirical techniques. First, it shows how to use a randomized experiment to elicit how much people value public services. Getting a good number for these valuations is important for public policy: You want to know how much people benefit when government expands access to a service, so you can weigh costs and benefits.

Second, it shows how these valuations can then be plugged into a discrete choice framework to simulate how people behave under counterfactuals you can’t simulate with an experiment (in this case, changing the nature of property rights in the community). This is really important. Focusing only on the parameters estimated by randomized experiments drastically limits the set of topics we can study. This paper shows how we can put those parameters to work (under reasonable assumptions) in answering questions outside of those addressed in the experiment. You can’t run an experiment changing people’s property rights, but the authors show you can do almost as well with good econometrics.

Abstract: In many societies, social norms create common property rights in natural resources, limiting incentives for private investment. This paper uses a randomized evaluation in Kenya to measure the health impacts of investments to improve source water quality through spring protection, estimate the value that households place on spring protection, and simulate the welfare impacts of alternative water property rights norms and institutions, including common property, freehold private property, and alternative “Lockean” property rights norms. We find that infrastructure investments reduce fecal contamination by 66% at naturally occurring springs, cutting child diarrhea by one quarter. While households increase their use of protected springs, travel-cost based revealed preference estimates of households’ valuations are only one-half stated preference valuations and are much smaller than levels implied by health planners’ typical valuations of child mortality, consistent with models in which the demand for health is highly income elastic. Simulations suggest that, at current income levels, private property norms would generate little additional investment while imposing large static costs due to spring owners’ local market power, but that private property norms might function better than common property at higher income levels. Alternative institutions, such as “modified Lockean” property rights, government investment or vouchers for improved water, could yield higher social welfare.

(Source: nber.org)