Forecasting the presidential election

Accurate forecasting of the outcome of the upcoming presidential election is important for a number of business reasons. Forecasting is an important part of financial planning and tax planning, for example.

Forecasting the outcome of the presidential race does not seem to be a difficult task at this point:

“Most nonpartisan political analysts currently see the electoral map tilting in the Democrats’ favor. Their ratings take into account a number of factors, from polls and political developments to a state’s voting history. Merging the ratings from four such handicappers—Cook Political Report, Rothenberg & Gonzales Political Report, Sabato’s Crystal Ball and Real Clear Politics—shows a lopsided consensus outlook if the election were held today, with Democrats winning 347 electoral votes to 191 for Republicans.” – WSJ 7/22/2016

How could the result of the election change from the analysts’ forecasts? What could change between now and November? The most likely way would be if the 4 closest margin states (Florida, Ohio, Virginia and Colorado) switched from Democratic to Republican, then Trump would win. A non-traditional but possible route for a Republican victory would be to gather such strong support from white blue collar people in Ohio, Pennsylvania, Michigan and Indiana to swing those electoral votes in his favor.

There is also a possibility that 3rd party candidate Gary Johnson, expected to peak at 20% of popular support, could actually surpass the two frontrunners.

Barring these unlikely events, a Democratic win seems likely and will be the basis of my business forecasts.