regression is statistically significant, providing evidence that the inclusion
of random county effects makes sense. At the same time, as can be seen,
the results remained remarkably stable. There is no change in the sign and
significance of the coefficients, despite the reduction in the magnitude of the
values. However, as stated earlier, this model relies on the lack of correlation
between the county random effects and the explanatory variables. Based on
Hausman, Hall & Griliches (1984), one can test this hypothesis by means
of an Hausman test that evaluates the random and the fixed effects estima-
tors. When applied to this setting, the test provides indirect evidence on
the correlation between the random effects and the explanatory variables.
The statistic equals 390.7 and thus we can not reject the null hypothesis at
the 1 percent level of significance.
Therefore, in a final specification, we use an alternative approach to deal
with the potential violation of the IIA assumption caused by the omission of
relevant variables. We estimate a CLM where we include a dummy variable
for each U.S. county. The estimation of this CLM is made by means of a
Poisson regression with fixed effects (see column 5, Table 3). The estimates
exhibit some noticeable changes. Agglomeration economies (both urbaniza-
tion and localization) are still significant and with the right sign. The same
is true for property taxes. However, the evidence on the significance of local
markets and the costs of land and labor disappears. A possible explanation
for the observed changes is that the estimates for this model are based ex-
clusively on time series variation. The time variability of our data may be
insufficient to identify the importance of these variables.
In sum, when controlling for ”county specific-effects” we find strong ev-
17