Explaining the Explanations of 100 Million Missing Women - Mechanism and Causality in Biology and Economics - page 231

there is a naturally occurring difference in the average sex ratio (country A, due to
the effect of HBV, on average has a higher sex ratio than country B) but on the
margin the sex ratio in both places can move (country B, due to its own local
concern, may have a higher sex ratio than it would normally have). In the case of
country B, we observe that there is a negative relation between HBV and sex ratio, a
result that is contradictory to what is described in the capacity claim of HBV (HBV
is a factor possessing the capacity to produce a positive change in sex ratio).
However, just as we have noted in the discussion of causal structure and capacity,
this paradoxical conclusion can be reconciled by resorting to a change in the
original structure. Our original simple causal structure contains only one causal
path: the HBV infection rate of country B leads directly to the sex ratio at birth of
country B. To accommodate the new fact that country B is a son-preference
country, we need to add a new causal path leading from country B's degree of
son preference to its sex ratio at birth. These two causal paths together constitute a
more complicated new causal model from which a new net result—a negative
association between HBV and the sex ratio—is generated.
7 The Causal Images of Mechanisms
We have identified two methodological approaches in the missing-women debate:
substitutive and complementary. We have also argued that the complementary
approach can be understood as an attempt to identify the causal structure. Studies
that seem to significantly conflict (such as Blumberg and Oster 2007 ) with the test
that is regarded as decisive (i.e., Lin and Luoh 2008 ) may be interpreted as attempts
to find the strengths of causal paths. Because it is the net result of the causal
structure that the empirical test tests against, the data cannot reject the existence
of the causal structure. All in all, there is an image of a causal structure in scientists'
minds.
Because of the influence of the received image of science, we often find that
empirical economists refer to correlation and regularity but seldom discuss mecha-
nism. In the sense of Machamer et al. ( 2000 ), we can say that the genuine underly-
ing structure that produces the phenomenon of the correlation between HBV and
sex ratio at birth—be it a positive or a negative relation—is the mechanism that
makes the relation what it is. When the manifest regularity between HBV and sex
ratio at birth is a positive relation, it is the mechanism that makes it so; when the
regularity is a negative one, it is still the mechanism that makes it so, but, this time,
the mechanism operates under a different structure that makes the originally
positive relation become negative. By ignoring the underlying mechanism, the
biological explanation is ruled out simply because a robust relation between HBV
and sex ratio at birth can no longer be found. However, a faint image of causal
structure—like Machamer, Darden, and Craver's mechanism schemata and
sketches of the underlying mechanisms—is still possessed by anyone who is
interested in the missing-women phenomenon. We can infer a great many possible