Returns to School Only Help if People Know About Them

JPAL has a nice write up a piece by Robert Jensen which looks at how perceived returns to education influence how much schooling a child receives. Jensen examines boys in the Dominican Republic and tests to see if providing accurate information on how much extra money a child will make by going to school influences the schooling decision.

The first question is how much extra does a child make by going to school. As Jensen notes there is no micro data on this available for the Dominican Republic so he does out and collects his own. He then finds the difference in wages between adults who attend secondary school and those that do not. Next he asks children about what they perceive to be the difference in wages between educated and non-educated workers. Their estimates are substantially lower than Jensen's estimates.

Here is the neat part he then created an experiment where half of the children are told the estimate returns and half receive no extra information. The idea is by telling children they will earn more if they go to school it will motivate them to attend more school. The experiment shows that children who receive the information go to about 1/5 to 1/3 of a year of school more over the next four years. Given how cheap the program is (just need to have seminars to tell kids about schooling) this is a good return.

Jensen does add some important caveats in his conclusion
"Of course, the desirability of such information-based programs will depend on the ability to provide accurate information on the returns to schooling, which may often be difficult. Further, even with accurate estimates, there may be reasons that the returns for the marginal child may not be as large as the currently measured average return"

The piece is published in the Quarterly Journal of Economics a top econ journal. It would be interesting to see this type of experiment replicated in other countries.

Could China Catch up to the US

China's economy grew at only 7.6% last quarter (China Daily). A rule of thumb tells us that if this rate of growth continues the economy will double in size roughly every 10 years. At $8,400 per capita GDP China's out put per person is roughly 1/5 of the United States. Doubling every 10 years would put China at $16,800 GDP per capita in 2022 and 33,600 in 2032 and given current growth rates it would then pass the US sometime before I retire.


Of course this assumes China will keep growing at its current rate. The main model of economic growth theory (the Solow model) predicts that growth rates slow down  once an economy is larger. This happens because each additional machine we have adds less than the previous one (for example last week I had no computer due to this computer failing, going from zero to one computer increased my productivity a lot, giving me a second computer when i have another that works doesn't increase my productivity that much.) This is also called diminishing returns.


Thinking about this made me recall something I read on a prediction by Paul Samuelson in the book Why Nations Fail the well known economist and textbook author predicted in 1973 that the USSR would have a GDP per capita equal to the US by somewhere between 1990 and 2015.

Brad Delong has an older post on how using the standard Solow model even if we assume that communism under China or the USSR has lower productivity (that is the same number of machines per person produces less) that a communist country could catch up to the US by essentially forcing savings and in the model any savings is used to buy more machines thus increasing output.

So we should watch not only China's growth but also changes in consumption and savings.

Finally, it is my duty as an economist to say it isn't necessarily bad if China has a higher GDP per capita than the US.

Maternal Literacy and Child Health

A new study released through UNICEF and other agencies examines the impact of the child support grant in South Africa. In short the program gives households below a certain income threshold a monthly cash payment around $30 a month.

One thing that caught my eye in the study is that the overall impact on children's height is not statistically different from zero. We care about how tall children are because several studies have shown in the long run children's height is strongly associated with future income also children who receive help or avoid shocks that stunt their growth tend to earn higher incomes when they grow up. This is consistent with a meta-analysis I did with co-authors that should be released shortly that cash transfers have small impacts on height for age.
The child support grant was found to improve height for age for children whose mother had at least 8 years of education, this is consistent with previous findings in Mexico (and I looked into this in a similar Nicaraguan program and found the same thing).

As the authors of the South African study suggest mothers may need to have schooling to fully implement the associated changes in behavior such as improved nutrition and hygiene that are needed to get the most out of the cash payments.

 h/t Roving Bandit 

Family Planning in Malawi

The below video on Family Planning in Malawi put out by the UK Dept for International Development (DFID) * and tweeted today by USAID demonstrates the main issues surrounding fertility in low income countries like Malawi. Economists see access to family planning and contraceptives as a supply and demand issue. On the supply side the question is can women get access to contraceptives are there clinics available. On the demand side do they actually want them notice in the video the clinics also includes education programs for the women about the benefits. In the video one woman also talks about how she asked her husband about getting birth control and then wanted to make sure he wouldn't pressure her for more babies. Intrahousehold issues and women's bargaining power are also key.

Fertility in Malawi has fallen substantially since 1980 when it peaked at 7.5 kids per woman falling to 6 kids per woman today (Gapminder). Nearly in every country over the last 30 years fertility has fallen, some may be surprised that countries like Brazil and India only have 1.8 and 2.6 kids per woman, respectively.

The video also demonstrates another issue as families have fewer kids (quantity is reduced) they can invest more in each one so (quality) increases.  One thing I teach my students is that in poorer countries children are in a sense a form of an old age pension or social security (since they will take care of their parents when the parents get older). During the last 30 years child mortality in Malawi has fallen from roughly 1 in 4 kids not making it to age 5 in 1980 to 1 in 9 today. This may also be related to falling birth rates


* in the past my research has received funding from DFID

Utility from Utilities: Principal Agent, Monopoly, and Pepco

Like many DC area customers of Pepco (our local power company) my house lost power last week* The Washington Post  in a recent story on Pepco claimed that the company under performs adjusting for the weather and tree cover that my area has. As someone who has lost power for multiple days at a time several times since moving into my house three years ago today I am sympathetic to this argument. So I put on my economist hard hat and got to thinking.

1. The Issue: a power company like Pepco is an agent that makes decisions about how much effort to put into making sure power is reliability provided. I am a principal who is effected by that decision. The Principal-Agent problem  can lead to many sub-optimal outcomes (no power is sub-optimal!), it also may lead to underinvestment if we can't observe Pepco's effort and random shocks occur. In my class a couple of weeks ago I showed my students a simple model where the probability of making a profit through farming was related to effort (which was costly). When someone worked on their own land they would worked twice as hard as if they worked on someone else who paid them only when a profit from a good harvest is realized. Furthermore, if the land owner paid the worker when a bad outcome occur it would only decrease effort.

2. Perhaps this suggests a solution. We should only pay Pepco when we have electricity (the good outcome)? Oh wait I guess we already do that. Perhaps Pepco's customers should be able to deduct from their bill a rate equal to the average power use from their house on the number of days they lost power (or perhaps some multiple of that the number). I would like to see an economic model of that

3. I don't know much about if Pepco is doing a good job. I also don't have the ability to choose another power company (unless I move). In this case Pepco has a monopoly. For some industries with high fixed costs a monopoly is natural (that's why we call it a natural monopoly), we wouldn't want 8 different companies stringer wires (well maybe we would want another at this point). When public utilities have a monopoly it seems logical to have outside experts evaluate if Pepco is doing a good job. I would like to read less about politicians saying people are fed up with power outages and more actual reports from impartial experts if Pepco is not properly behaving in terms of its natural monopoly.

update Arnold King at AEI has similar thoughts as a PEPCO customer h/t to Dan Rothschild

* I should add I was in San Francisco for all but 4 hours of the power outage, but my wife and daughter were there and they are part of my household utility function which includes utility from utilities.

Hilton Starbucks Thinks I Won't Walk 1 block for $1

I have shared most of my thoughts from the Western Economics meetings, but one final thought after I returned home. The Hilton at Union Square in San Francisco has a Starbucks that charges $2.85 for a cup of coffee the Starbucks just down the street charges $1.85 (or something there about), so the Hilton charges a $1 more.

Why might this be the case? First, hotel guests might not know there is a starbucks one block away (I discovered this only because the Hilton had no free wi-fi (see this old post) and a 15 minute line for coffee). Second hotel guests are often attending conferences and may be time constrained (which is why I did buy one cup of coffee for $2.85). Third, someone else might be paying for the coffee purchase (in my case that wasn't true since I'm not using Towson travel funds for coffee).

Evidence Economists Influence the NBA but not Safeway

A recent paper by Price and Wolfers showed that in the NBA " that more personal fouls are awarded against players when they are officiated by an opposite-race officiating crew than when officiated by an own-race refereeing crew. " The paper got quite a bit of press and another paper showed you could make money betting on teams that had a higher percentage of players the same race as the referees. So did the paper have any effect on the NBA. First the NBA claimed their referees weren't bias and that they didn't do anything to address the issue since there wasn't an issue. Now a new paper presented at the Westerns Economics Association Meetings with Pope, Price and Wolfers shows that "that racial bias continued to exist in the NBA prior to the media coverage, but then completely  disappeared during the four years thereafter" So not proof that the NBA listened to Price and Wolfers, but it is consistent that they did.


Now this morning I was out trying to buy ice, because my house hasn't had power since Friday (it's Tuesday). I go to 7-11, the gas station and Safeway and all are sold out of bags of ice. The worst was Safeway still had signs up advertising a sale on ice. Typically economists suggest when demand for a product rises due to outside forces (like say ones that knock out power for 1 million people) ice prices should go up. Some people call this gouging. In fact many did in Silver Spring. I say raise the price so people only buy the ice that is necessary. 


I also did ice arbitrage. I filled up large soda cups with ice that only cost 25 cents and brought them home instead of a bag of ice. Of course now I've got the Power (back)