Pennsylvania’s risk-limiting election testing, explained.

Pennsylvania is one of 15 states that use this type of test to track midterm election results.

  • Sam Darklau

Workers sort mail-in ballots at the Lehigh County Government Center in Allentown on November 8, 2022.

Matt Smith / For Spotlight PA

Workers sort mail-in ballots at the Lehigh County Government Center in Allentown on November 8, 2022.

As part of what is routinely done to review Pennsylvania’s election results, the State Department has asked counties to conduct what it calls a risk-limiting review.

The State Department tested this type of ballot inspection in 2020 when it ordered counties to inspect 45,000 randomly selected ballots. The exam this year showed evidence that the results were correct — despite efforts by some Republican leaders to do so amplify former President Donald Trump’s voter fraud lie.

Election scrutiny became apparent after the 2020 election, when activist groups began calling for them, believing they exposed widespread inadequacy. This year’s risk mitigation review and another routine review of two percent of ballots in all districts showed the results were correct. Numerous other studies experts from both parties confirmed this finding.

This year, for the first time, the State Department reviews ballots with a statewide risk-limiting audit. From now on there will be one every year. Acting Secretary of State Leigh Chapman has called them the “gold standard of robust election testing.”

Fourteen other states, including swing states like Georgia and Michigan, use risk-limiting audits to review their election results.

What is a Risk Limiting Audit or RLA?

They are designed for it reduce the likelihood of an election being confirmed for the wrong candidate. To that end, election officials selected one race for review — in this case, the gubernatorial race between Attorney General Josh Shapiro and Senator Doug Mastriano (R-Franklin) — and randomly selected groups of ballots Thursday for each county to review.

A group of agency employees rolled ten-sided dice to generate a 20-digit “seed number” that tells a computer software program which ballot groups counties need to review.

“Review teams in each county will manually review and hand-count the votes cast for Pennsylvania governor from the ballots in those groups,” Assistant Secretary Jonathan Marks said. “Once all counties have reported the results of their audit, the State Department will collate and evaluate the statewide results to determine if the risk limit has been met.”

In other words, whether the result matches what the audit found.

Marks added that since this year’s exam will only review the gubernatorial race, each county only needs to review a relatively small number of ballots.

“The number of ballots required by the RLA is breed-specific and depends on the difference between the winning candidate and the runner-up. The smaller the margin, the more ballots have to be checked.”

Shapiro won the race by more than 14 percentage pointsbut an audit has yet to be conducted to ensure the result is accurate statewide.

Why should an RLA be preferred over any other type of assessment?

Any choice, counties check at least two percent of all ballots checked for errors. At least 2,000 are checked in some smaller counties. But that small sample of ballots isn’t always the most representative of a county’s election result — and examining that group alone isn’t always the most effective way to weed out potential problems.

Pennsylvania’s electoral system generally produces accurate election results, but stray voting errors or voting record mismatches do occur from time to time.

A risk mitigating audit aims to identify and investigate such issues to determine if they are widespread.

“If the reported result is deemed incorrect, an RLA will continue to escalate and may result in a full contest recount,” added Marks.

In other words, if districts find problem ballots in the first set of ballots they check, they check another set to see if they find any more. More mistakes mean more scrutiny, designed to stop bad election results before it’s too late.

Risk-limiting audits work particularly well in tight races: poll workers have to check more ballots from the start, increasing the likelihood of spotting counting errors.