Category: Econometric models

Jay DeSart updated his long-range model with April polling data. The most recent forecast slightly favors Trump in a likely match-up against Clinton. In particular, the DeSart model predicts Trump to gain 50.9% of the vote, compared to 49.1% for Clinton. Trump is also predicted to win the electoral college with 295 votes, compared to 243 for Clinton.

The vote-share forecast currently enters the combined PollyVote with a weight of 3%. In comparison, today’s PollyVote forecast predicts the Democrats to gain 53.3% of the vote.

Jay DeSart updated his long-range model with February polling data. The most recent forecast favors the Republican candidates in hypothetical match-ups against both Clinton and Sanders.

For the likely outcome of a Clinton-Trump race, the DeSart model predicts Trump to gain 51.6% of the vote, compared to 48.4% for Clinton. This forecast currently enters the combined PollyVote with a weight of 4%. Trump is also predicted to win the electoral college with 311 votes, compared to 227 for Clinton.

The forecast is in contrast to the most recent PollyVote forecast, which predicts the Democrats to gain 52.2% of the vote.

Bruno and Véronique Jérôme updated their state-by-state political economy model. The model now predicts the Democrats to gain 49.2% of the two-party vote, compared to 50.8% for the Republicans. The model forecast currently enters the combined PollyVote with a weight of four percent.

Interestingly, although the model predicts the Democrats to lose the popular vote, they are still expected to win the electoral vote. The latest forecast predicts the Democrats to gain 325 electoral votes, compared to 213 for the Republicans.

Helmut Norpoth provided an update of his Primary model forecast. The initial forecast posted on February 27 relied on a polling data, not actual results, for the Democratic primary in South Carolina. Now that the results are in, which give Hillary Clinton a much bigger lead than was indicated by the polls, the forecasts for the various general election match-ups are as follows:

Revised general election forecasts of the Primary model
Race	Democratic vote	Republican vote	Probability of winning
Clinton-Trump	47.5	52.5	87%
Clinton-Rubio	52.4	47.6	86%
Sanders-Trump	42.3	57.7	99%
Sanders-Rubio	47.2	52.8	89%

At the moment, there are forecasts from five component methods, and five econometric models, available. Therefore, the primary model enters the combined PollyVote forecast with a current weight of 4%. Not surprisingly then, the PollyVote forecast is largely unaffected by this change and still predicts the Democrats to gain around 52.0% of the two-party vote.

Polly added the Primary model by Helmut Norpoth to its econometric models component. The model is based on state electoral histories and candidates’ performance in early primaries.

Assuming a Clinton-Trump race, the model predicts a clear victory for the Republicans, with Donald Trump gaining 54.7% of the two-party vote (compared to 45.3% for Clinton). This is the largest lead for the Republicans in any forecast available to date. In other words, the forecast is an outlier. In comparison, today’s combined PollyVote forecast predicts the Democrats to gain 51.9% of the vote, compared to 48.1% for the Republicans.

Polly added the long-range model by Jay DeSart to its econometric models component. The model is based on state electoral histories, national polling data as well as two variables that attempt to estimate next year’s election context.

Assuming a Clinton-Trump race, the model’s latest forecast predicts a close outcome, with a slight advantage for the Republicans. That is, the model predicts the Democrats to gain 49.3% of the national two-party vote, compared to 50.7% for the Republicans. In comparison, the latest PollyVote forecast forecasts a vote share of 52.0% for the Democrats and 48.0% for the Republicans.

Note: There was a minor mistake in calculating the combined econometric model forecast as a result of adding the DeSart model on February 12th. The effect on the combined PollyVote forecast was negligible, however. The error was corrected on February 15th.