Bride price and the returns to education for women

Continuing the development economics theme this week, one of the most important and most studied development projects is the massive school building program in Indonesia in the 1970s, the Sekolah Dasar INPRES. Over 61,000 primary schools were constructed - about one for every 500 children aged 5-14, which roughly doubled the number of primary schools in the country, and increased the enrolment rates of children aged 7 to 12 from 69 percent in 1973 to 83 percent in 1978.

The effects of this program on educational attainment and wages for men were evaluated by Esther Duflo in this AER paper (ungated here). She found:

The INPRES program led to an increase in educational attainment in Indonesia. On average, the estimates indicate that the program led to an increase of 0.25 to 0.40 years of education (0.12 to 0.19 years for each new school built per 1,000 children), and increased by 12 percent the probability that an affected child would complete primary school. The estimates also suggest that the program led to an increase of 3 to 5.4 percent in wages.

A subsequent NBER working paper by Lucia Breierova and Esther Duflo (ungated here) showed that the effects on women's education were much smaller. Both studies make use of the fact that date and region of birth determine each person's exposure to the increased schooling (which overcomes the identification problem that would usually arise - since family characteristics determine the choice of schooling, and subsequent labour market outcomes).

Which brings me to this new paper (PDF) by Navaa Ashraf, Natalie Bau, and Nathan Nunn (all from Harvard) and Alessandra Voena (University of Chicago). Using the same dataset and identification, they set out to investigate the effect of bride price on education attainment among women. Bride price is the custom of the groom's family making a payment to the wife's family at the time of the marriage (it is the opposite of dowry). It's not about buying a bride - it is typically interpreted as rewarding the parents of the bride for their years of investment in their daughter, and compensating them for her no longer contributing to their household (directly). Bride price is (still) a reasonably common practice in parts of Africa and Asia.

The paper's findings are interesting:

We show that among ethnic groups that practice bride price, the amount that the bride’s family receives as a bride price payment increases with the level of education of the bride. Completing primary school is associated with a 100% increase in the bride price payment, completing junior secondary is associated with a further 40% increase, and completing college with another 100% increase. These relationships are very robust and remain strong even when conditioning on a large set of observable characteristics, as well as potentially endogenous characteristics like the groom’s education.

Essentially, they find that the impact of the school building program in Indonesia was much greater for ethnic groups that practice bride price, and was virtually zero for other ethnic groups. Moreover, they show similar results for Zambia (which had a similar, albeit smaller, wide-scale school building program in the 1990s and early 2000s). The reason why bride price had such positive effects on educational attainment for girls is that:

...bride price provides a greater incentive for parents to invest in girls' education, and it is these parents that are more likely to take advantage of the increased supply of schools by educating their daughters.

Bride price often gets an unfair rap, but this is one situation where it appears to have had a positive impact - young girls would have had less schooling in Indonesia had it not been for the existence of the bride price custom.

[HT: Kevin Grier at Cherokee Gothic]

El Nino, climate change, and agriculture in the tropics

Following on from yesterday's post on developing countries, today I read an interesting article from the AER Papers and Proceedings issue earlier this year (ungated version here; PDF), by Solomon Hsiang (University of California, Berkeley) and Kyle Meng (University of California, Santa Barbara). In the article, the authors investigate the relationship between the El Nino Southern Oscillation (ENSO) and agriculture in tropical and temperate countries. ENSO has been shown to cause large and systematic changes in local climatic conditions around the world, but these effects have the greatest impact in tropical countries.

The authors use as a measure of ENSO an index of sea surface temperature in the eastern Pacific Ocean (a commonly used measure). They find:

that ENSO systematically affects country level temperature and rainfall in the tropics... A rise in the ENSO index by +1°C increases local temperatures in the tropics by +0.27°C and lowers rainfall by -4.6 cm on average (combined over two years). For temperate countries, temperatures actually fall due largely to changes in atmospheric and ocean circulations, but only by half as much, and there is a small but insignificant positive effect on rainfall.

More importantly though, they find:

A +1°C increase in the ENSO index lowers cereal yields -2%, total cereal production -3.5%, and agricultural income -1.8% on average across the tropics. These effects are highly statistically significant and suggest that rises in prices do not fully compensate countries for declines in agricultural output... crop yields increase in temperate countries when the tropical Pacific warms, albeit with a smaller magnitude that is less significant.

Given that some (but not all) research suggests that future climate change will be linked to stronger ENSO events (for a great review see here), these results suggest that poor countries (which are disproportionately located in the tropics) could face decreased agricultural output, decreased food security, and decreased economic growth in the future. Perhaps it also helps to explain in part the poor past growth trajectories of these countries as well.

How love conquered marriage

One of the key differences between developed and developing countries is the presence (or absence) of complete and functional markets. In my graduate development economics class, we spend a fair amount of time talking about the causes and implications of incomplete labour markets, credit markets, and insurance markets. The absence of functional insurance markets is a big problem for people in developing countries and as you would expect, they have developed various strategies to insure themselves (either individually, as a household/family, or as a community) against adverse shocks such as bad weather. These strategies include share-cropping (which shares the risk of weather shocks between the tenant farmer and the landowner), choosing low-yield but low-risk crop varieties, temporary rural-urban migration to diversity sources of income, or diversifying geographically through marriage of family members into families located in other villages (or regions, or even countries).

One of the ways that a family can ensure that they diversify geographically is for parents to arrange marriages for their children. Late last year, Marginal Revolution pointed me to this interesting paper (PDF) by Gabriela Rubio (University of California, Merced), and I've been meaning to write about it for a while. Rubio uses data from the Indonesian Family Life Survey (IFLS) to investigate whether improved agricultural outcomes have reduced the need for arranged-marriage-as-insurance, and have led to the observed decline in arranged marriage in Indonesia (and in many other countries, though Rubio notes that India, Pakistan and Bangladesh are exceptions). From the paper:

The goal of this paper is to understand the main driver(s) of the transition by proposing and testing empirically a model of marital choices. I first show that this transition away from arranged marriages in favor of self-choice or “love” marriages is correlated with increases in education, formal employment, urbanization, and declines in agriculture. These trends are common in all the countries where micro-data is available, suggesting that despite having different institutions at work, there is a fundamental economic explanation behind these changes in marriage institutions.

Based on these patterns, I build a simple model of marriage choice. I assume that arranged marriages serve as a form of informal insurance (as suggested in the literature of sociology, anthropology and economics, e.g. Rosenzweig and Stark (1989)) whereas other marriages (outside one’s networks) do not...

The model predicts that arranged marriages disappear when the net benefits of the insurance arrangement decrease relative to the (unconstrained) returns outside of the social network. When this is the case, parents invest in more education for the child, effectively increasing her outside option and, thus, the probability that she will reject the arranged marriage.

Rubio first demonstrates that arranged marriage acts as a form of (imperfect) insurance. She then uses the progressive introduction of the Green Revolution across Indonesia from the 1960s to the 1980s as a quasi-experiment, to test her theory about the impact on arranged marriage. She finds:

...that the Green Revolution increased returns to schooling by an additional 2.1% to 4.7% per additional year of schooling, it increased mean income of agricultural households, and importantly, it decreased their income variance by 8.1% and 8.3%, respectively.

Importantly, because the income variance decreased, that made insurance less necessary, which in turn led to the main results:

As predicted, the Green Revolution resulted in a decline of 9 to 20 percentage points in the probability of having an arranged marriages [sic] for the cohort exposed to the Green Revolution, and in an increase in education of 0.3 to 0.5 years of schooling for the same cohort.

The results are robust to various alternative specifications that Rubio tries. Which may lead us to conclude that the Green Revolution conquered arranged marriage, leaving love to take over.

However, it does leave one important question unanswered - what is so different about India, Pakistan and Bangladesh? They've had similar increases in agricultural incomes to Indonesia (and other countries that Rubio identifies as having declining rates of arranged marriage), and they've had a reduction in variability of incomes as well due to the Green Revolution. So the need for arranged-marriage-as-insurance is lower in these countries too, and yet arranged marriage persists. Are South Asians simply more intent on maintaining marriage traditions than Indonesians, do they use arranged marriage to signal their group affiliations and maintain social capital, or are they much more risk averse (such that insurance is still necessary even in the wake of improved incomes)? These are questions to be answered in future research.

A collection of the best bits from "Dear Undercover Economist"

I've been reading through the back catalogue of Tim Harford books. Earlier this year I read "Dear Undercover Economist", which is a collection of Dear Economist columns Harford wrote for the Financial Times. Tim Harford is a great author (and a great speaker - he was a keynote speaker at the New Zealand Association of Economists conference a few years back), and many of these columns (written in the style of an 'agony aunt' column) are hilarious.

This book is well worth reading, although since most of these columns are available online (either at the Financial Times (which may require you to sign up or subscribe) or on Harford's own site), I've collected a few of my favourites here (the short titles in the bullet points are mine):

• Do the costs outweigh the benefits of getting a girlfriend?
• Should children get sex education at school?
• Peltzman effects and driving like a lunatic
• Should you go along with an arranged marriage?
• Explaining the credit crunch if it was a fairy tale
• An answer to those pesky Nigerian scammers
• Where do missing socks go?

Enjoy!