All Academic, Inc. Research Logo

Info/CitationFAQResearchAll Academic Inc.
Document

"What's In It For Me?": Why Members of Congress Pursue Oversight
Unformatted Document Text:  Having shown that the coefficient estimates for all but one of the covariates are statisticallysignificant and in the hypothesized direction, I now turn to the substantive interpreta-tion of these coefficients. The interpretation of coefficient estimates from a model thatemploys a negative binomial probability distribution is not intuitive. I have thereforeincluded Table 4, below, which reports the expected change in the number of committeeor subcommittee oversight hearings when one shifts each covariate, in turn, from its 25 th percentile value to its 75 th percentile. For instance, the simulated first differences for the seniority variable can be interpreted as indicated that one could expect a subcommitteewhose members are in the 75 th percentile in terms of seniority to hold 0.51 fewer hearings than would be expected of a subcommittee in the 25 th percentile for member seniority. Table 4: First Differences in the Expected Number of Subcomm. Oversight Hearings variable from (25 th percentile) → to (75 th ) first diff. 95% C.I. Divided Gov’t unified −→ divided +0.87 (0.24, 1.49) Oversight Mandate no −→ yes +5.32 (3.51, 7.37) Margin of Victory 19.2% −→ 49.1% +0.02 (-2.06, 1.94) Bills Passed 1.54 −→ 2.48 -0.27 (-0.51, -0.04) Dim. 1 Pref. Heterog. 0.69 −→ 0.81 +0.51 (0.09, 0.93) Dim. 2 Pref. Heterog. 0.39 −→ 0.55 +0.47 (0.02, 0.92) Chamber Seniority 2.91 −→ 4.44 -0.51 (-1.00, -0.04) Rules Comm. Deference 20.0% −→ 66.7% -0.71 (-1.15, -0.26) Simulated first differences in the expected number of subcommittee oversight hearings, when one shiftseach given explanatory variable from its 25 th to its 75 th percertile value, in turn, while fixing all continuous variables at their means, interval variables at their medians, and nominal variables at their modes (dividedgov’t at “divided,” and oversight mandate at “no”). Quantities of interest estimated using negativebinomial regression in Zelig. (cf. Imai, Kosuke, Gary King, and Olivia Lau. 2007. “negbin: NegativeBinomial Regression for Event Count Dependent Variables,” in Kosuke Imai, Gary King, and Olivia Lau,“Zelig: Everyone’s Statistical Software,” http://gking.harvard.edu/zelig.) Given that the mean number of hearings held by individual committees and subcom- mittees during this period was 3.54 (with a right-skewed standard deviation of 4.89),the first differences associated with divided government, oversight mandate, bills passed,preference heterogeneity, seniority and Rules Committee deference are sizeable. 65 4 Conclusion The theme of oversight as a “second-best” activity is apparent in virtually all of the find-ings presented in this paper. An analysis of subcommittee transfers in Section 2 showedthat MCs tend not to prefer seats on oversight-focused subcommittees. This finding held 65 These results are even more acute when one performs a simulated first differences analysis as above, but instead shifts some of these interval and continuous variables, in turn, from their minima to theirmaxima. For example, shifting bills passed from its minimum to its maximum value is associated with2.6 fewer oversight hearings in expectation. Similar shifts in 1st dim. preference heterogeneity, 2nd dim.preference heterogeneity, seniority, and Rules Comm. deference yield +4.6, +2.7, -2.5, and -1.6 changes,respectively (all of which achieve conventionally accepted levels of statistical significance). 38

Authors: Feinstein, Brian.
first   previous   Page 38 of 46   next   last



background image
Having shown that the coefficient estimates for all but one of the covariates are statistically
significant and in the hypothesized direction, I now turn to the substantive interpreta-
tion of these coefficients. The interpretation of coefficient estimates from a model that
employs a negative binomial probability distribution is not intuitive. I have therefore
included Table 4, below, which reports the expected change in the number of committee
or subcommittee oversight hearings when one shifts each covariate, in turn, from its 25
th
percentile value to its 75
th
percentile. For instance, the simulated first differences for the
seniority variable can be interpreted as indicated that one could expect a subcommittee
whose members are in the 75
th
percentile in terms of seniority to hold 0.51 fewer hearings
than would be expected of a subcommittee in the 25
th
percentile for member seniority.
Table 4: First Differences in the Expected Number of Subcomm. Oversight Hearings
variable
from (25
th
percentile) → to (75
th
) first diff.
95% C.I.
Divided Gov’t
unified −→ divided
+0.87
(0.24, 1.49)
Oversight Mandate
no −→ yes
+5.32
(3.51, 7.37)
Margin of Victory
19.2% −→ 49.1%
+0.02
(-2.06, 1.94)
Bills Passed
1.54 −→ 2.48
-0.27
(-0.51, -0.04)
Dim. 1 Pref. Heterog.
0.69 −→ 0.81
+0.51
(0.09, 0.93)
Dim. 2 Pref. Heterog.
0.39 −→ 0.55
+0.47
(0.02, 0.92)
Chamber Seniority
2.91 −→ 4.44
-0.51
(-1.00, -0.04)
Rules Comm. Deference
20.0% −→ 66.7%
-0.71
(-1.15, -0.26)
Simulated first differences in the expected number of subcommittee oversight hearings, when one shifts
each given explanatory variable from its 25
th
to its 75
th
percertile value, in turn, while fixing all continuous
variables at their means, interval variables at their medians, and nominal variables at their modes (divided
gov’t at “divided,” and oversight mandate at “no”). Quantities of interest estimated using negative
binomial regression in Zelig. (cf. Imai, Kosuke, Gary King, and Olivia Lau. 2007. “negbin: Negative
Binomial Regression for Event Count Dependent Variables,” in Kosuke Imai, Gary King, and Olivia Lau,
“Zelig: Everyone’s Statistical Software,” http://gking.harvard.edu/zelig.)
Given that the mean number of hearings held by individual committees and subcom-
mittees during this period was 3.54 (with a right-skewed standard deviation of 4.89),
the first differences associated with divided government, oversight mandate, bills passed,
preference heterogeneity, seniority and Rules Committee deference are sizeable.
65
4
Conclusion
The theme of oversight as a “second-best” activity is apparent in virtually all of the find-
ings presented in this paper. An analysis of subcommittee transfers in Section 2 showed
that MCs tend not to prefer seats on oversight-focused subcommittees. This finding held
65
These results are even more acute when one performs a simulated first differences analysis as above,
but instead shifts some of these interval and continuous variables, in turn, from their minima to their
maxima. For example, shifting bills passed from its minimum to its maximum value is associated with
2.6 fewer oversight hearings in expectation. Similar shifts in 1st dim. preference heterogeneity, 2nd dim.
preference heterogeneity, seniority, and Rules Comm. deference yield +4.6, +2.7, -2.5, and -1.6 changes,
respectively (all of which achieve conventionally accepted levels of statistical significance).
38


Convention
All Academic Convention is the premier solution for your association's abstract management solutions needs.
Submission - Custom fields, multiple submission types, tracks, audio visual, multiple upload formats, automatic conversion to pdf.
Review - Peer Review, Bulk reviewer assignment, bulk emails, ranking, z-score statistics, and multiple worksheets!
Reports - Many standard and custom reports generated while you wait. Print programs with participant indexes, event grids, and more!
Scheduling - Flexible and convenient grid scheduling within rooms and buildings. Conflict checking and advanced filtering.
Communication - Bulk email tools to help your administrators send reminders and responses. Use form letters, a message center, and much more!
Management - Search tools, duplicate people management, editing tools, submission transfers, many tools to manage a variety of conference management headaches!
Click here for more information.

first   previous   Page 38 of 46   next   last

©2012 All Academic, Inc.