SmartShot Logo

How would the 2016 election look if electoral votes were allocated based on current population?

Matt Cassidy and Bryan Whitaker

Right after election day 2016, we started getting questions about how the Electoral College might have been different had the Census and redistricting happened before this election. The idea being that our population is changing so quickly demographically and our people are so mobile, perhaps the allocation of Electors isn’t keeping up with the rate of change in our population distribution.

So we took note, and kept an eye out for the Census Bureau’s Population Estimates Program (PEP) update which just came out on December 20, 2016.

Using the PEP data, we found not so encouraging news for Democrats and Progressives — Hillary Clinton still would have lost. Put another way, there would not be massive enough shifts in the Electoral college if reapportionment were to occur based on the new Census estimates.


Florida, North Carolina, Oregon and Texas would all gain one seat, while Illinois, Michigan, Minnesota and Pennsylvania would each lose 1 seat.  This equates into a net change of only one 2016 Electoral Vote, with Trump taking 307 to Clinton’s 231.


Using this same PEP data to look forward to what this may mean for the US House of Representatives down the road, New York would get the 435th seat and Rhode Island the 434th seat, both just narrowly avoiding losing a seat.  Texas and Colorado came in 436th and 437th, both likewise just narrowly missing out on adding a seat.  Meanwhile Minnesota and Michigan were 438th and 440th – both not too far away from not losing a seat.  Also Montana would have taken the 439th seat, making it by far the closest state to leaving “At-Large” congressional status.

We’ve already started thinking ahead about what the trends in our population shifts may mean for the actual 2020 Census, redistricting and the 2020 Presidential Election. It’s important to keep in mind, the reapportionment results from the April 2020 Census will not be released until December of 2020 – after the 2020 election and will not go into effect until the 2022 midterms and ultimately the 2024 Presidential election.  The moral being that Democrats cannot depend on shifting demographics and will have to win back a lot of 2016 states in 2020, regardless.

That said, we took the state level estimates so far for the decade – 2010-2016 – and applied linear forecasting to them, to forecast 2017, 2018, 2019 and 2020.  The result was different, though only incrementally better.   Under this scenario, Michigan, Minnesota, Illinois and Pennsylvania all lose one seat each.  These go to California which adds 2 and Arizona and Colorado which add 1 each.  This changes the Clinton-Trump outcome to 233 to 305. In the US House of Representatives, Washington takes the 435th seat holding on to its current total.  Illinois, Michigan and Pennsylvania all just miss out on holding on to their seat, coming in with rankings of 436th, 437th and 439th.  North Carolina and Texas are 438th and 440th, just missing out on adding an additional seat.



ABOUT MATT CASSIDY | Senior Data Analyst, TargetSmart
Matt Cassidy spent over six years at the National Committee for Effective Congress (NCEC) where he drew the map of New Jersey legislative districts that will be in use through the year 2021. He also aided in state legislative redistricting efforts in the states of Florida, Michigan, Missouri and Vermont as well as Congressional redistricting in New Jersey. For more on Matt, you can read his full bio here.


ABOUT BRYAN WHITAKER | Chief Innovation Officer, TargetSmart | @whitakerb2
Bryan is a leader in non-profit information and communications technology management striving to use data, analytics and innovative software to change the world. Bryan is the former COO of NGP VAN and was the Director of Technology at the DNC during President Obama’s re-election campaign. For more on Bryan, you can read his full bio here.