Last time I went to the grocery in Lexington, one of the stock clerks began tailing me. He held his distance at first, keeping me in eyesight but not approaching, until I started browsing an out-of-the-way clearance rack. Then he made his move.

“You’re that UK professor, right?” He asked the question furtively, as though we were two spies meeting under flickering lamplight. “Stephen Voss?”

I admitted my guilt.

“Let me ask you something.” He said it under his breath, and didn’t wait for my nod. “Is Andy Beshear going to win the governor’s race?”

Sigh. After such suspenseful buildup, his query was a letdown. Nine times out of ten, it’s the first one I get after someone learns I’m a political scientist. “Who’s going to win the next election?” National journalists who’ve been calling me for the last few months to discuss the governor’s race approach the topic in fancier ways, but usually what they want to know boils down to the same horse-race question: “Is a Democrat really going to win reelection in Kentucky?”

Of course, if being a political scientist meant I’d been issued a crystal ball along with my diploma, I might be able to satisfy these interrogators. But that’s not how science works.

Science can pinpoint a pattern. For example, pre-election polls usually underestimate GOP support in Kentucky, often by large margins (partly for the same reasons they underestimated the Republican winner’s strength in Louisiana last week). That’s a possible source of hope for Attorney General Daniel Cameron, who has consistently trailed Beshear. Polls badly missed a Republican gubernatorial victory here in 2015, so maybe they’re doing it again.

Science can document changes. For example, white professionals have abandoned the Republican Party in the years since Trump’s victory and the fall of Roe v. Wade. That’s good news for Gov.? Beshear, who needs a strong showing in the affluent Golden Triangle region of the state to overwhelm a lopsided Republican tilt almost everywhere else.

Science can seek explanations. For example, the timing of the Democratic Party collapse in Appalachia leads me to attribute it, more than anything, to stringent energy regulations enacted by President Obama, which his opponents successfully vilified as a War on Coal. That’s decent news for Beshear, because Kentucky Democrats have been able to avoid the full force of coal country’s rejection of the national Democratic Party (as Beshear himself did in 2019).

Science can help us understand processes. For example, although election reformers usually treat the decision whether to vote as personal, based on how individuals balance the costs and benefits of turning out, political scientists have long known that participation is heavily influenced by organizations – such as parties or interest groups – that invest resources to whip up citizen engagement. That’s great news for Beshear, who (unlike the resource-deprived Democrat crushed in Louisiana) has a massive amount of cash available to mobilize supporters.

Science can puncture myths. Two examples spring to mind. One is the myth that undecided voters break for the challenger. Another is the myth that low turnout helps Republicans. Both would be good news for Cameron, except research doesn’t support either rule of thumb. Low turnout likely helps Democrats in Kentucky, especially when their candidate enjoys high name recognition and an enthusiastic base – so Cameron is the one who should have his fingers crossed that Kentuckians will show up for the off-year election.

All the things science permits — identifying patterns, documenting changes, testing explanations for why the world behaves as it does, puncturing myths – can be useful when issuing forecasts. But knowing how the world generally works doesn’t allow an analyst to say with certainty how it will behave in a single instance, leading to wishy-washy predictions that confuse and frustrate laypeople.

I might estimate Republican chances of taking the governor’s mansion this year to be worse than a coin toss, given my application of research to the current context, but I won’t be betting the house on it. If your meteorologist tells you there’s a 45% chance of rain, that doesn’t mean it won’t rain. It means bring an umbrella.

Another thing about scientific research frustrating to laypeople is that sometimes it can even support conclusions that end up being wrong – especially when probing new developments, such as the spread of an unknown virus.

We have such a disease infecting the body politic now: an especially virulent form of partisanship that has spread from political leaders to the citizenry over the last 30 years. Political scientists are still scrambling to catch up with this political polarization, as new data on the contagion keep rolling in.

A recent snapshot of the fevers roiling American politics comes from Project Home Fire, based at the University of Virginia. Their survey released last week documented so many symptoms of democratic breakdown among both Democrats and Republicans — willingness to suppress civil liberties, endorsement of violence against opponents, even an openness to splitting the country — that I wouldn’t know how to pick highlights. The nation is sick.

One pessimistic conclusion to emerge from scientific research about this polarization is that officeholders no longer have much hope of keeping their jobs by converting voters. The incumbency advantage, as it’s called, has eroded away because Americans are so driven by their party identification – especially their hostility toward the other party – that trying to win us over by doing a good job is useless.

But the great thing about science is that, even when wrong, it can fix itself based on new information. This year’s gubernatorial election should force political scientists to reconsider how much leaders can benefit from trying to please skeptical voters, even those who initially might seem deplorable.

Beshear started the campaign season, and he heads into the Nov. 7 election, with some of the highest job-approval ratings of any governor in the nation. Should he retain his post – heck, even if he narrowly loses among such a pro-Trump electorate – he’ll at least have shown that Kentucky voters are more persuadable than outsiders have been assuming.