The DeSart model provided an updated forecast of the election result. According to this model, Clinton will obtain 50.4% of the two-party vote share in New Hampshire, whereas Trump will end up with 49.6%. In comparison, on August 14, Clinton was predicted to achieve only 49.6% of the vote.
In New Hampshire, the popular vote is usually decided by a narrow margin. This is the reason why the state is commonly viewed as a swing state, which makes it particularly interesting from a forecasting perspective.
Putting the results in context
Single econometric models should be interpreted with caution, since they can include substantial biases. Instead of relying on results from single econometric models, one should look at combined econometric models or, even better, a combined forecast that uses different methods and data.
Results in comparison to other econometric models
If we look at an average of New Hampshire econometric models, Clinton's current two-party vote share is at 52.2%. Compared to her numbers in the DeSart model Clinton's econometric model average is 1.8 percentage points better.
The DeSart model compared with PollyVote's forecast
The results of the DeSart model for Clinton are thus 3.2 percentage points below the combined PollyVote, which at the moment predicts a value of 53.6% in New Hampshire. In comparison, a look at the PollyVote national prediction for Clinton shows that the actual results are 2.6 percentage points lower.