Many reputable national polls conducted over the past two weeks show former Vice President Joe Biden with a lead of about 10 percent over President Donald Trump. A gap of this magnitude is larger than the margin of polling error, so a popular vote victory for the Biden-Harris ticket is highly likely. However, as we learned in 2000 and 2016, the electoral vote, not the popular vote, will determine who is inaugurated president on Jan. 20.
In 2016, Trump won states with 306 electoral votes, 36 more than the 270 needed for election. When projecting this year’s electoral vote, it is useful to think in terms of a range of scenarios, rather than a single likely outcome.
At this time, with nearly two weeks to go until the end of voting, the most probable, but by no means definitive, scenario would have Biden winning states or districts with 48 electoral votes that Trump won in 2016: Michigan, Wisconsin, Pennsylvania and the single electoral votes in the 2nd district of Maine and the 2nd district of Nebraska. This combination would give Biden 280 electoral votes, 10 more than the minimum needed.
The most uncertain state in this scenario is Pennsylvania. Although polls show Biden leading there by an average of nearly 7 points, both the Biden and the Trump campaigns see Pennsylvania as closer than either Michigan and Wisconsin. The two campaigns are devoting a disproportionate share of candidate time and advertising resources to the Keystone State in the final weeks of the campaign. Results in Pennsylvania may not be known for several days because the state courts are allowing ballots postmarked by Nov. 3 and received by Nov. 6 to be counted, a decision that the Supreme Court let stand on a 4-4 tie vote.
A narrow Biden electoral win is currently the most probable outcome. However, one cannot set aside the possibility of Trump winning the electoral vote by winning Pennsylvania by a small margin. With the other results noted above, a Trump win in Pennsylvania would give him 278 electoral votes.
To prevent this outcome, the Biden campaign is working to expand the playing field to states that Trump won in 2016 but that are more competitive in 2020. The first group of states in this category includes Arizona, North Carolina and Florida. 
Recent polls show Biden with a small lead — within the margin of error — in all three states. All three of these states have been undergoing substantial demographic change in the past decade, becoming younger and more diverse, with an increasing share of Latino voters. These developments tend to advantage Democratic candidates. To overcome a potential loss in Pennsylvania, Biden would need to win either Florida, or both Arizona and North Carolina.
Finally, there are three states where polls indicate the presidential contest is basically tied — Iowa, Georgia and Ohio. These states have been reliably Republican in recent elections, but the Biden campaign believes it is competitive in all of them, because of demographic change (Georgia), dissatisfaction with the state of the rural economy (Iowa), or a return of working-class voters to traditional Democratic voting patterns (Ohio).
In conclusion, I believe the most likely outcome of the presidential election will be a Biden popular vote margin over Trump in the mid-to-high single-digit range, and Biden winning somewhere between 280 and 300 electoral votes. 
However, there is a not insignificant chance of two other outcomes: a Trump electoral vote win by holding on to enough of the states he won in 2016 to overcome a larger popular vote deficit than four years ago, or a substantial Biden win, with a double-digit popular vote lead, and an electoral vote total in the 320 to 350 range.  
Whatever the outcome, turnout is likely to be very high, possibly exceeding the modern-era high of 62.8 percent of the voting-age population in 1960, and perhaps even coming close to 65 percent.
Eric L. Davis is professor emeritus of political science at Middlebury College.