Once we control for all these factors, where did Ron Paul do best? One factor we have not been
able to control for is the strength or efficacy of the Ron Paul organization. Presumably this is the
“omitted variable” that explains most of the remaining variation across states. We can figure out where
Ron Paul’s organization must have been strongest by looking at those states that significantly
underperform or overperform the baseline model. The mechanics of this operation are as follows. First,
we substitute our estimated coefficients into the regression equation, thus:
Then we substitute for each of the independent variables [[caucus, turnout, clinched, candidates)
the particular value for each state. The resulting ln[[Paul vote) figure is the predicted value of the log of
Ron Paul’s vote share for each state. We can subtract this predicted value from the actual value, and the
difference is the amount by which Ron Paul outperformed what we would expect, given primary or
caucus format, turnout, timing [[whether McCain had clinched or not), and number of candidates on the
ballot, all factors outside Paul’s control. High values of [[actual vote – predicted vote) indicate that Paul
did better in those states than we would expect; low values mean he did worse. Therefore, high values
presumably indicate that Paul had a better, more effective organization.
Table II ranks the states by the number of log points by which Paul outperformed his predicted
performance in each state.
Table II: Paul’s Unexplained Performance by State
Overperformance
0.7993741
0.7624688
0.4739113
0.4645284
0.3769193
0.3725045
0.3609405
0.3455322
0.296046
0.2901108
0.2458186
0.2377231
0.2362665
0.1911311
0.1876662
State
New Hampshire
Idaho
South Dakota
Washington
North Dakota
Pennsylvania
Michigan
Montana
Oregon
Vermont
District of
Columbia
Tennessee
Maryland
New York
Maine
0.1718122
0.1714756
0.1538725
0.1459794
0.123376
0.075878
0.0525227
0.0177805
0.0128961
0.001461
-0.0221498
-0.0315268
-0.0460235
-0.0604255
-0.0773115
-0.1129332
-0.1456997
-0.1672701
-0.1790915
-0.231692
-0.2629383
-0.2664576
-0.2991312
-0.3194487
-0.3745465
-0.4054149
-0.4389548
-0.444923
-0.4810518
-0.5319622
-0.6164439
-1.052599
California
New Mexico
Alaska
Nebraska
Rhode Island
Illinois
Missouri
Minnesota
Nevada
Arkansas
New Jersey
Ohio
Wisconsin
Texas
Virginia
Arizona
South Carolina
Delaware
Iowa
Oklahoma
Connecticut
Florida
Kansas
Indiana
Georgia
North Carolina
Utah
Kentucky
Alabama
Massachusetts
Colorado
Mississippi
So we can now answer the question: In which state did Ron Paul’s organization do best? The
answer is New Hampshire. New Hampshire was the most pro-Ron Paul state, although Idaho ran a close
second. The other states are well behind. At the other end of the spectrum, Paul did really badly in
Mississippi. We can only deduce that voters and activists in Mississippi strongly disliked or ignored Dr.
Paul’s antiwar, libertarian message. Colorado, Massachusetts, Alabama, Kentucky, Utah, and North
Carolina were also pretty poor states for him.
Bookmarks