Request Username
Can't sign in? Forgot your username?
Enter your email address below and we will send you your username
INTRODUCTION
Sheldon Jacobson
Professor
University of Illinois at Urbana-Champaign
Email: [email protected]
Advanced analytics has found a home in the political arena. From forecasting elections, to allocating advertising budgets for campaigns, to computational redistricting to combat gerrymandering, advanced analytics have demonstrated an uncanny knack for gaining insights where the plethora of data makes it challenging for the mind to grasp. Numerous websites are available (e.g., fivethirtyeight.com and electionanalytics.cs.illinois.edu) to interpret the data and present it in a manner that people can understand and appreciate.
Election forecasting is at the core of these activities. The availability of polling data provides an excellent analytical opportunity to hone prediction skills, consider turnout scenarios, and examine political trends. Analyzing and presenting the data allows voters to understand how elections may unfold on a national scale. Access to this information may impact how they vote and even whether they vote at all.
When we first launched the Election Analytics Editor’s Cut in 2016, this was before the unexpected (by many) victory of Donald Trump over Hilary Clinton in the 2016 U.S. Presidential Election. This victory, considered to be beyond the grasp of analytics, was comparable to the 1948 upset victory of Harry Truman over Thomas Dewey (click here for a detailed analysis). However, as with all modeling efforts, data drives the analysis, and this was no exception.
The effectiveness of any candidate or political party also relies on state redistricting. To combat gerrymandering, one can use discrete optimization methods, census data, and past election outcomes to produce district plans that limit partisan bias. Groups that focus on computational redistricting (such as http://redistricting.cs.illinois.edu) aim to make these methods more efficient and offer a transparent approach to the political process.
This updated Editor’s Cut looks at a collection of articles on election forecasting, advertising analysis, computational redistricting, and what we call, Analytics in Action, all providing a glimpse into how advanced analytics is playing a growing role in politics. Moreover, with the 2020 decennial census providing a fresh set of data to redraw congressional and state districts, and algorithmic advances providing new ways to employ artificial intelligence and operations research, election analytics will gain even more traction in the future.
Election forecasting is at the core of these activities. The availability of polling data provides an excellent analytical opportunity to hone prediction skills, consider turnout scenarios, and examine political trends. Analyzing and presenting the data allows voters to understand how elections may unfold on a national scale. Access to this information may impact how they vote and even whether they vote at all.
When we first launched the Election Analytics Editor’s Cut in 2016, this was before the unexpected (by many) victory of Donald Trump over Hilary Clinton in the 2016 U.S. Presidential Election. This victory, considered to be beyond the grasp of analytics, was comparable to the 1948 upset victory of Harry Truman over Thomas Dewey (click here for a detailed analysis). However, as with all modeling efforts, data drives the analysis, and this was no exception.
The effectiveness of any candidate or political party also relies on state redistricting. To combat gerrymandering, one can use discrete optimization methods, census data, and past election outcomes to produce district plans that limit partisan bias. Groups that focus on computational redistricting (such as http://redistricting.cs.illinois.edu) aim to make these methods more efficient and offer a transparent approach to the political process.
This updated Editor’s Cut looks at a collection of articles on election forecasting, advertising analysis, computational redistricting, and what we call, Analytics in Action, all providing a glimpse into how advanced analytics is playing a growing role in politics. Moreover, with the 2020 decennial census providing a fresh set of data to redraw congressional and state districts, and algorithmic advances providing new ways to employ artificial intelligence and operations research, election analytics will gain even more traction in the future.