Our Data Plans for the 2018 Election
Among the many things we are doing at the MIT Election Lab, gathering, cleaning, and disseminating data are activities that never stop. We plan to make available to the public a variety of datasets that will provide context and clarity to the election results as they come in. These datasets are part of the public goods we provide to the election science community. We hope they will be widely used (with attribution, of course!).
I detail our plans for this effort below.
For starters, to help provide social context to the returns, we are producing a series of county-level demographic datasets using data from a variety of official sources — mostly the U.S. Census Bureau. The idea is that these data can be downloaded and easily merged with election returns, once available. These datasets will include measures of educational attainment, employment, income, nativity, population, rurality, and sex.
To provide political context, we will also be including some historical election returns, for those who wish to analyze vote shifts by county, district, state, etc.
Also in the works are a series of datasets centered around elections laws and policies. These include measures of early voting, election day registration, voter ID, vote by mail, online registration, and a series of online lookup tools for voters. Several of these will also give careful readers a potential early look at some of the components for the next release of the Elections Performance Index.
When it comes to the 2018 midterm election returns, we will provide election night results in an easily usable form from all states that make them available. We expect that this may take a day or two after the election, but that should not stop anyone with earlier access to returns from merging in the contextual data we will already have available. These data will likely be pushed out in several iterations so please check back often.
To find our data, the quickest place to access it as it becomes available will be our GitHub page for the 2018 election. We will eventually post the certified election returns on our Dataverse and website, but the quickest turnaround will always be on GitHub. The contextual data will be posted the day before the election, and we will start posting available returns as we get them.
We encourage free and open use of our data (although we ask that you mention that it came from MEDSL), and please always let us know if you have ideas for how we can improve. You can file issues in GitHub or get in touch directly. If you are less interested in data and more interested in findings, please keep an eye on our various social media outlets (as well as this Medium page) to see our own analyses of the election.
Cameron Wimpy is the Research Director at the MIT Election Data & Science Lab.
MEDSL is dedicated to applying scientific principles to how elections are studied and administered, with the aim of improving the democratic experience for all U.S. voters. Follow updates on Twitter and sign up for our quarterly newsletter.
