How The Inquirer is reporting the results of Pennsylvania’s 2022 midterm elections

Pennsylvania elections have gotten a lot more complicated.

The dramatic expansion of mail voting over the last two years has brought major changes to how voters cast ballots, how votes are counted, and how election results are reported.

Still, The Inquirer’s mission remains the same: to provide the most accurate and up-to-date information as possible. That includes letting you know when a race has been called by The Associated Press; providing the latest vote counts, along with contextual information to help understand those numbers; explaining the vote count as it occurs; and covering any developments that arise.

Here are some key things to know about how we’re reporting the results of Tuesday’s midterm elections, which includes the U.S. Senate race between Lt. Gov. John Fetterman, the Democrat, and celebrity doctor Mehmet Oz, the Republican; the governor’s race between state Attorney General Shapiro, the Democrat, and State Sen. Doug Mastriano; U.S. House seats, including several hotly contested ones; state House and Senate races that will help determine control of the legislature; and more.

The Inquirer’s results come directly from the Associated Press, which has workers across the state calling in county election results as they come in.

Elections are run by counties, which then report results to the Pennsylvania Department of State throughout the night. The state displays those results on its unofficial election returns page, but that feed often lags behind the real-time vote count in many counties. Philadelphia, for example, reports its vote totals on its own page, which is usually far ahead of the state’s compilation.

» READ MORE: Pennsylvania 2022 midterm election results

The AP’s network means those county-level results flow in throughout the night much faster than what you’ll see on the state’s website.

The Inquirer itself doesn’t call races — instead we rely on the AP, which has a long track record of accurately declaring winners in even the closest of elections.

Those race calls aren’t a projection about a likely outcome. The AP only declares a winner when it’s clear a candidate has won.

“If our race callers cannot definitively say a candidate has won, we do not engage in speculation,” the AP says in its guide to its process.

It’s important to remember that race calls are unofficial declarations from news organizations. That doesn’t make them unreliable — the AP was 100% accurate in calling the presidential and congressional races in every state in 2020. It just means the final official result takes time.

Votes always take time to count, which is why Pennsylvania’s vote certification doesn’t occur until 20 days after election day.

Counties are better prepared, and have fewer mail ballots, than in 2020, and the vote count should be significantly faster. Most votes cast this election should be counted within a few hours of polls closing. That’s partly because most votes will be cast in person, with those results reported on election night. Most counties will begin counting mail ballots at 7 a.m., when state law allows them to start, and run their vote counts around the clock until they’re done.

But a court ruling will slow the process by requiring counties to check the handwritten dates on the outside of mail ballot envelopes. And once the bulk of the votes are counted, there will always be more to count in the days and weeks ahead.

There are will be a number of paper ballots that have to be manually adjudicated, with county elections boards deciding what to do with them. In the past, for example, officials have had to decide what to do with ballots that voters returned to the wrong county.

There are also always some votes that need to be counted by hand because of issues with the ballots themselves, such as having stray marks or coffee stains on them. There are provisional ballots, too, which counties can’t start counting until Friday.

And overseas and military ballots can arrive up to one week after Election Day if they’re postmarked in time.

Anytime we present election results, you’ll see a percentage of expected votes that have been reported. That gives a sense of how far along in the vote count we are.

The AP’s expected vote counts are estimates that change throughout the night. They become more accurate as votes are counted, giving a clearer and clearer picture of how much is left.

It also means the numbers will shift, including going up or down as the expected vote count is adjusted. That doesn’t mean something nefarious is going on — it just means the AP is updating its estimate to reflect new information. If the percentage of expected votes counted suddenly goes down, for example, all that means is the AP has increased its estimated vote total.

We use that expected vote number to present each candidate’s vote share. So in the first moments after polls close, if only 1% of the expected votes are in, we won’t say that a candidate has won 50% of the vote so far — that would be misleading. Instead, we’ll tell you the candidate has won 0.5% of the expected vote so far.

One thing you won’t see: the term “precincts reporting.” In the past, when 95% or so of Pennsylvania’s votes were cast in person, knowing how many precincts had reported their in-person results was a handy way of knowing where the count stood. Now that a significant portion of votes are cast by mail, we’re using the expected vote count instead.

Geography is very important in politics. In addition to ongoing urban-rural partisan divides, Pennsylvania is a state made up of many distinct regions, each with unique political, cultural, social, and economic histories.

That means it’s important to know where votes are coming from as they come in throughout the night. Democratic voters in Philadelphia, for example, may have very different preferences from Democrats in Southwestern Pennsylvania. Republicans in the Philadelphia suburbs might have different preferences from those in Northeastern Pennsylvania.

We’ve divided Pennsylvania’s 67 counties into eight regions. They’re not perfect — any attempt to define the regions is surely setting us up for hate mail — but they’re a useful rough geographic categorization for understanding where the results are coming from.

Here are our regions: