Explaining the Explanations of 100 Million Missing Women - Mechanism and Causality in Biology and Economics - page 220

3 How Economists Explain 100 Million Missing Women
Blumberg's hepatitis B hypothesis is interesting, but no economist took it seriously
until Emily Oster's ( 2005 ) article was published. Oster tested Blumberg's idea on a
population level using both time-series and cross-sectional data. In particular, Oster
used the historical data of Taiwan's universal vaccination of all newborns begin-
ning in 1984 to conduct an empirical test to see whether there is indeed a significant
positive correlation between the variable of HBV prevalence and an increase in the
sex ratio at birth. Taiwan's vaccination case forms a natural experiment that is ideal
for testing the correlation, because economists can analyze the historical data of
births to vaccinated and unvaccinated mothers to see whether there is a great gap in
offspring sex ratio between these two groups; if there is indeed a significant
difference in sex ratio at birth, then HBV has a positive effect on the percentage
of children who are male; otherwise, it does not.
Although Oster's testing result showed that there is a significant correlation
between HBV prevalence and sex ratio at birth, she noted a caveat to her conclu-
sion. Oster observed that Taiwan's vaccination program coincided quite closely
with an increased availability of fetal sex-determination technology and a probable
increase in sex-selective abortion. Consequently, the magnitude of the difference in
sex ratio at birth between vaccinated and unvaccinated mothers is likely to be
smaller than it otherwise would be, because the effect of vaccination on decreasing
sex ratio at birth is offset by the countervailing behavior of adopting fetal sex-
determination technology, which was motivated by the cultural cause of son
preference.
However, in a cross-country analysis, Oster, by applying the least-square method
on single equations, found a significant correlation between HBV and sex ratio at
birth. She also stated that about 75 % of Sen's 100 million missing women could be
explained by parental infection with HBV, implying that son preference plays a
lesser role in explaining the missing women. Soon after, Blumberg ( 2006a , b )
characterized Oster's finding as one of the great achievements of HBV research,
because it supported his previous findings and confirmed that HBV does have
effects on gender.
In response to Oster's biological explanation, Monica Das Gupta, a supporter of
the cultural explanation, maintained that the Chinese sex ratio at birth for the first
birth was always within the normal range of 1.05-1.06. A higher sex ratio at birth
was observed only in subsequent births. In addition to these two empirical testing
results, Das Gupta found that an extremely high sex ratio at birth was observed
mainly among women who had previously given birth only to daughters. Together,
these results strongly suggest that it is son preference, rather than HBV, that has the
significant effect on distorting sex ratio at birth and therefore it is the former, rather
than the latter, that is the cause of the missing women (Das Gupta 2005 , 2006 ). In
response, Oster ( 2006 ) pointed out that she does not disagree with the cultural
explanation; rather, she opposes Das Gupta's position that “the support for cultural
explanations allows to conclude that the biological explanation is not particularly