|Dear Readers: This is the latest in a series of political science forecasts for the 2016 races for the White House and Congress. We’ll be featuring forecasts from nine different individuals and/or teams this year, which James E. Campbell is assembling as part of a project for PS: Political Science and Politics that we are also featuring in the Crystal Ball. These models are based on factors such as the state of the economy, polling, whether an incumbent president is running for reelection, and other indicators. They can often be a better predictor of the eventual results than polls alone, and many are finalized months before the election.
We are pleased to feature the work of the many top political scientists who have built these models, both in an attempt to predict the outcome of the election and, more importantly, to identify the factors that actually affect presidential and congressional races. Below, Campbell lays out the details and outlook of his presidential forecasting model, and Michael S. Lewis-Beck and Charles Tien present their Senate forecast. We have also updated our tables to include all the presidential and congressional forecasting data presented in this series.
— The Editors
The Trial-Heat Presidential Election Forecasting Models: The Convention Bump Model
By James E. Campbell of the University at Buffalo, SUNY
There are two companion presidential election forecasting models based on the preference polls and election-year economic growth. The two models are the Labor Day Trial-heat and Economy model and the Convention Bump and Economy model. The models predict the national two-party vote percentage for the in-party presidential candidate. The models are essentially sophisticated readings of the preference polls, taking into account the historical adjustments made for the competitive fall campaign, the economy, and incumbency. Both models have been estimated using data for elections from 1948 to 2012, excluding 2008 because the Wall Street Meltdown of financial institutions violated the models’ assumption that no unanticipated cataclysmic event intervenes between the forecast and the election.
It is still too early to obtain a forecast from the Labor Day model, but the necessary information to generate a 2016 forecast from the Convention Bump and Economy model is now available. The Convention Bump model was first developed and used in 2004. It was developed in response to the parties holding later conventions, possibly contaminating poll readings for the Labor Day model, but also to provide additional data leverage for the forecasts. Setting aside the 2008 Wall Street Meltdown election, the convention bump and economy has been quite accurate, missing the actual in-party vote percentages by less than two percentage points in both 2004 and 2012. The median out-of-sample error for the model is 1.3 percentage points.
The Convention Bump and Economy model is constructed from three predictors. The first is the two-party percentage of support for the in-party’s presidential candidate prior to the first national nominating convention. The second is the change in that support from before the first convention to after the second convention. This is the net convention bump for the in-party candidate. The third predictor is the second quarter annualized change in the real GDP. Original data was used in the estimation when it was available and growth rate reading is from the Bureau of Economic Analysis’s late August (second estimate). Because the 1980 decline in GDP was so extreme that it exerted undue leverage on estimated effects, it has been capped at a 3.5% decline (still far below that in any other modern presidential election). Also, in order to take into account the differences between incumbent and successor in-party candidates, the GDP data are centered around a 2.5% growth rate and then halved for successor candidates.
The 2016 election posed one problem for these trial-heat models. The trial-heat data used in previous elections had been collected by the Gallup Organization. They had collected trial-heat data in presidential elections since 1936. Gallup, however, has not reported trial-heat data to this point in 2016. In lieu of the Gallup trial-heat data, I have used the median of the seven national polls reported on RealClearPolitics closest to the date at which the Gallup reading would have been taken.
The Forecast: Based on 52.2% of the two-party pre-convention preference polls (a 47% Democratic to 43% Republican split), a 0.4% net convention bump favoring the Democrat, and a weak 1.1% real GDP growth in the second quarter of 2016, the Convention Bump and Economy model’s forecast is that Hillary Clinton will receive 51.2% of the two-party national popular vote to 48.8% for Donald Trump. The forecast was made on Aug. 26, 2016.
Uncertainty: Based on the distributions of out-of-sample errors for previous elections, the Convention Bump and Economy’s probability of correctly forecasting that Clinton will receive more than 50% of the national popular two-party vote is 75%.
Political Economy Model (Senate)
By Michael S. Lewis-Beck of the University of Iowa and Charles Tien of Hunter College, CUNY
Description: The model is similar to the House model with the addition of a short-term institutional advantage that measures how many seats the president’s party has up for re-election. The structural model reads as:
Predictors: 1) Political Popularity = the job approval rating for the president in the June Gallup Poll, 2) Economic Conditions = growth rate of real disposable income over the first two quarters of the election year, 3) Midterm Status = 0 for presidential election years and = 1 for midterm election years, 4) Seats Exposed = number of seats the president’s party has up for reelection.
The Forecast: A four-seat gain for the Democratic Party, made on July 29, 2016
Uncertainty Estimate: Percentage likelihood that the forecast has predicted the Senate majority party winner is 62% (based on within-sample estimates across the series).