Legislative Bargaining and the Dynamics of Public Investment

This corrigendum provides details of corrections to Tables 5, 6, and 7 from our recent article (Battaglini, Nunnari, and Palfrey 2012). In the course of some followup work with our data, we discovered that the estimates for the U sessions reported in Tables 5 and 7 are incorrect, as were several entries in Table 6. The only notable change is in the significance level of the constant terms in the last two columns of Table 5. These two corrected constant terms were reported in the article as being insignifcant. They are signficantly different from zero. The claim made in the last sentence of Finding 6 (p. 420, lines 5–8) is still supported by the corrected constant term estimates from columns (2) and (4) of Table 5. The claim is not supported by the constant term estimates reported in columns (1) and (3), but those two constrained models are both rejected by the unconstrained models of columns (2) and (4) at the 1% and 5% significance levels, respectively. The estimates of both constant terms from the more elaborate model in Table 7 also support the claim. The signs and signficance levels of all other coefficients in both tables are unaffected. A few other minor errors have also been corrected. The constant term in column 1 of Table 5 was mislabeled in the article as significant at the 10% level, when it was significant at the 5% level. Column (4) of Table 5 was mislabeled as (5). The corrections to Tables 6 are all minor. Cells with corrected entries are highlighted in gray.

TABLE 5. Logit estimates. Dependent var: Pr {vote = yes}. SE clustered by subject in parentheses; *significant at 10% level; **significant at 5% level; ***significant at 1% level

TABLE 6. Proposal Acceptance Rates, Theory vs. Experiments. “% Acc as Pr” is the percentage of accepted proposal that theory predicts to pass and “% Rej as Pr” is the percentage of refused proposals that theory predicts to be refused. The number of observations is in parentheses.

TABLE 7. Logit estimates. Dependent var: Pr {vote = yes}. SE clustered by subject in parentheses; *significant at 10% level; **significant at 5% level; ***significant at 1% level.