Elections, Audits, and an Alpaca

By Eliza Steffen
B.A. Political Science, 2020
Elections & Redistricting Intern,
National Conference of State Legislatures

This summer, I am working in the Elections & Redistricting team at the National Conference of State Legislatures (NCSL). NCSL is a nonpartisan organization that works to improve the effectiveness of all 50 state legislatures, from California to New Hampshire, and advocate for their role in the federal government. We provide research and analysis on a wide variety of policy issues, and link legislators and experts to each other. My team focuses on election and redistricting policy, from the nitty gritty of election administration technology to the recent U.S. Supreme Court case (Gill v. Whitford) on partisan redistricting.

There are two big projects I’m working on for much of my time here:

• Compiling Candidate Demographics
  First, Hannah Zimmerman (the other Stanford intern, in the Legislative Strengthening department) and I are compiling candidate demographics – age, gender, occupation, and race/ethnicity for all the 6,066 state legislative seats up for election this November.
• Comparing Initiative and Popular Referendum Processes Across 24 States
  Second, one of the staff and I are creating a new, comprehensive resource comparing the initiative and popular referendum processes across the 24 states that allow them. A popular or citizen referendum, sometimes known as the “people’s veto” is when voters can petition for a popular vote to repeal a law passed by their state legislature. Filing a ballot measure petition, and collecting the requisite number of signatures collected can be a complex and idiosyncratic process, so we’re excited to create a way to compare the different processes more easily.

Conducting a risk-limiting audit requires a lot of complex statistics, but the main thing you need to know is that it uses algorithms to determine how many ballots need to be audited based on the closeness of the race.

On July 6th, my co-workers and I had the opportunity to attend Denver County’s portion of Colorado’s second ever post-election risk-limiting audit (RLA). They’re the first state in the U.S. to actually conduct one, but recently a few other states have codified it (Rhode Island) or want to. RLAs are more or less a newer, more accurate way to figure out whether election results are accurate by using more advanced technology- code, etc. Conducting a risk-limiting audit requires a lot of complex statistics, but the main thing you need to know is that it uses algorithms to determine how many ballots need to be audited based on the closeness of the race. RLAs tend to take up less time and are more efficient – excellent for county budgets and state election administrators alike. Here, I rolled a die to determine one of the “random seed” numbers used to decide which ballots to physically audit.

I’ve also had the opportunity to write for NCSL’s blog, mainly this article on the June 26th primaries. So far, 30 out of 50 states have had their primary elections, meaning we are over halfway through primary season. July is more or less “summer break” primaries, so we won’t have another until Tennessee on August 2nd. At the moment, I’m working on an article looking at a more historical view of state elections from an interview with Lou Cannon, the former Washington Post D.C. reporter and Ronald Reagan biographer. It was fascinating to talk to someone with such a long and esteemed career in political journalism, and I learned a lot about how we should be thinking about the November midterms.

And in between all of this, there’s still time for fun – my first week featured a visit from an alpaca.

Read more at the Out West Student Blog »