I Know What Is Going to Happen Tonight

A Guide to What to Expect as America Votes Today

Eight years ago, I got sort of famous.

In early March of 2016, I published a long-form analysis of that year’s election in response to a simple question: “Can Donald Trump beat Hillary Clinton in November?”

At the time, very few people thought so. Trump hadn’t even fully locked up the Republican nomination yet, and most pundits dismissed him as a little more than a joke candidate. No one seemed to take his candidacy very seriously.

Good analysis is dispassionate and separated from one’s own emotional investment or bias, which is one of the big things I think a lot of people got wrong in 2016. They couldn’t ultimately see past their own opinion. Despite my distaste for populism and my misgivings about Trump the candidate, I saw the political landscape differently, and said so.

As I looked at the fundamentals of the race, I saw the beginnings of a political realignment taking place, as the populist wing of the Republican Party ascended with Trump at its head. I saw some fundamental problems with polls that caused me to be skeptical of their results. I also saw shifting demographic characteristics that I felt were going to be important in 2016.

I tried very hard to be objective, and in the end this is what my prediction boiled down to:

Even with all of those states [Nevada, Colorado, Iowa, New Mexico, and Virginia] going to Clinton, there is a huge problem she faces in rust belt states like Pennsylvania and Michigan, which have a tremendous number of white, blue collar, working class voters who blame free trade, and countries like China and Mexico for their job losses and economic decline.

This is also true of Maine's 2nd District, which is only worth a single electoral vote, but can easily be won by Trump for the same reasons.

[…]

And that doesn't even touch the possibility -- which is real -- that Trump could win a state like Wisconsin, which would lead to a 20 vote swing in his direction.

The only state I got wrong was Iowa, which I incorrectly believed would go to Clinton due to it having gone to the Democrats in three out of the prior four election cycles.

When the November election arrived and Trump — against virtually all expectations and predictions — won the race, a lot of people sought to explain what just happened, which is why my March prediction attracted so much notice.

Every election since I made that prediction, I end up getting asked (often by media pundits) to look into my crystal ball and predict what will happen this year. Given that today is election day 2024, I figured I might talk a bit about what I see when I look at the election this year.

Before I do, though, let me get one thing clear: everything you are about to read should be treated very skeptically in your mind. I don’t have a crystal ball — no one does — and my ability to predict races is directly related to the information I get, and how I analyze and interpret it.

It is true, I've spent countless hours dissecting data, parsing through polling errors, and scrutinizing early voting patterns. I've analyzed demographic shifts, pored over state-by-state dynamics, and considered every conceivable variable that I thought could influence the outcome.

But in the end, I got 2016 right because I noticed several things early on that many people missed. There is every possibility, nay likelihood, that this year I am the guy missing something and everything I think ends up being wrong. So take this all with a big grain of salt.

As long as you’re okay with that, keep reading. I’ll get to my actual predictions in a moment, but first I want to talk about the things that make predicting this race so hard.

Polling: As Fundamentally Broken as Ever

If there's one lesson we've learned from recent elections, it's that polls are not to be trusted.

I want to be clear about my position on this, though. The reason that polls have been such a disaster is not because pollsters are biased, or engaging in political advocacy to try to manipulate the outcome. They are trying, and trying very hard, to be right about the races they survey because their entire business relies on being able to correctly identify public sentiment.

It is just that polling is so very difficult to get right. Much more than it used to be.

Rather famously, in 2016 and 2020, pollsters underestimated Donald Trump's support, leading to widespread shock when the results rolled in. There were many explanations for why the failures happened, but the leading contender according to most analysts was some version of the so-called “nonresponse bias,” whereby Trump supporters were less likely to answer pollsters, causing pollster models to miss sampling the somewhat invisible voters.

Making things more confusing, polls were mostly dead on in both 2018 and 2022. So if you’re keeping track at home, you would score polls “terrible” in 2016, “great” in 2018, “terrible” in 2020 and “great” in 2022.

Fast forward to 2024 and it would only be natural to ask: will they be terrible again, or great?

No one knows of course, but the pollsters seem to think that they’ve figured things out — much like they did in 2020 — and that another polling error is unlikely.

And yet…

Basic Structural Problems in Polling

Despite advancements in technology and data analytics, polls are becoming less reliable, even when they prove to be accurate. Put another way, it is very hard to have any faith in them today, because of the repeated catastrophic failures intermixed with positive results.

Accurate polling requires sampling the correct universe of voters to capture true intentions. Pollsters build models based on assumptions — educated guesses — about who will vote. Over the past decade, plummeting response rates have intensified this challenge. With people increasingly unwilling to participate, pollsters rely more on complex modeling. Online polls attempt to mitigate this but introduce new biases and limitations.

When pollsters misidentify the voter universe, disaster ensues. This is exactly what happened in both 2016 and 2020, as pollsters believed they knew what the electorate was going to look like, but ended up being very wrong. Despite their adjustments after 2016, pollsters simply were not able to capture the complexities of voter behavior.

Pollsters base models on previous elections — who turned out and for whom they voted — assuming historical patterns will repeat. However, we are in the midst of a volatile political realignment, with radical shifts in voter alignment with certain parties and candidates, which make past outcomes less useful in building present day expectations. Voter loyalties are less predictable as traditional party affiliations are weakening, and this has been particularly noticeable in elections that feature Donald Trump at the top of the GOP ticket.

Certain demographics are also chronically underrepresented in polls. Rural voters, individuals with lower educational attainment, and some minority groups are less likely to respond, leading to skewed samples. Technological changes compound these issues. The decline of landlines and rise of cell-phone-only households make it harder to reach representative samples.

The COVID-19 pandemic added complexity as there were radical changes to both voting methods — particularly the shift to mail-in ballots and early voting — and voter motivation. Models based on previous elections couldn't account for these unprecedented changes.

Finally, the biggest challenge — nonresponse bias — skews results, as certain groups end up being far less likely to participate in polling, leading to samples that don't reflect the electorate.

Pollsters have been trying to find ways to adjust to this, and respond to make their polls more accurate, but we simply don’t know whether or not they were successful this year, and we won’t know until after the votes are counted.

Given systemic issues affecting accuracy, skepticism is warranted. While polls can indicate general trends, the complexities of modern electorates require new approaches to understanding voter behavior.

The problem is, what pollsters are doing to adjust may be making things worse.

Poll Herding

One of the things we have seen emerge this year as a problem is a phenomenon known as “herding.” Statistician and election forecaster Nate Silver has extensively written about this issue, and is very concerned about the effect it may be having on public opinion surveys. Here’s how it works.

Pollsters, wary of being the outlier, produce results that suspiciously converge around a tight race, with many swing state polls showing the candidates within a mere two-point margin. This happens when pollsters refuse to release polls that are outliers to the consensus polling average (thereby reinforcing that average), for fear of getting egg on their face if they are wrong. It can also occur when pollsters adjust their polling methodology to make the consensus outcomes more likely.

Silver argues that the problem is present in all of the swing states, but is most pronounced in Wisconsin:

The problems are most acute in Wisconsin, where there have been major polling errors in the past and pollsters seem terrified of going out on a limb. There, 33 of 36 polls — more than 90 percent — have had the race within 2.5 points. In theory, there’s just a 1 in 2.8 million chance that so many polls would show the Badger State so close.

Eli McKown-Dawson delved into this phenomenon as well recently, highlighting that the herding effect creates an illusion of consensus that may not actually reflect reality. Think of the probabilistic dynamics at play like this: if you were to flip a coin seven times, the odds of it landing heads each time are about 0.78 percent, and yet that is essentially what we're seeing with polls consistently showing near ties in battleground states.

The problem here is that the pollsters have become so risk averse, with none of them wanting to appear to get it wrong this year, that they may in fact be corrupting their own numbers to the point where they are largely useless. Many analysts who understand this problem are suggesting — something that I believe — that the outcome may not actually end up being that close.

Demographic Realignment in America

Demographics aren't destiny, but they certainly shape the battlefield on which political contests are fought. This election is no different, with several key demographic shifts that could tip the scales having become apparent. As these shifts continue to happen, it becomes harder to understand where people belong.

Education and Class

The education divide among white voters has become one of the most significant factors in American politics. White voters without a college degree have increasingly gravitated toward the Republican Party, a trend that has only intensified with Trump's populist movement fully transforming the GOP.

In contrast, white voters with college degrees, especially women, have been moving toward the Democrats. This shift is particularly pronounced in suburban areas, which could be crucial in swing states like Pennsylvania, Michigan, and Wisconsin.

This divide isn't just about education, it is also about cultural identity and perceived values of social groups. The Democratic Party's focus on progressive social issues has tended to resonate with college-educated voters but has at the same time alienated those without degrees who feel that they have been treated with contempt and disregard by a cultural elite, made up of those very same college-educated people.

Today, more than ever, the establishment and the working class believe that they share virtually nothing in common with the other, and as that perception has increased the tribal partisan re-sorting has only exacerbated itself, making predictions very difficult.

The Changing Politics of Race

African American voters have been a reliable base for the Democratic Party for more than a century, viewed as an essential part of the liberal coalition since at least 1932. After the Civil Rights battles of the 1960s, however, black support for the Democratic Party went from strong to nearly uniform.

In 1960, Richard Nixon won 32 percent of the black vote against John F. Kennedy. Twenty years later, when Ronald Reagan won a historic landslide victory against incumbent Jimmy Carter, he won only 14 percent of the black vote. By the time Barack Obama defeated John McCain in 2008, the Republican share of the African American vote was only 4 percent.

The reasons for this are varied, but ultimately trace to the decision by Lyndon Johnson to push the Civil Rights Act through Congress, and the successful Republican strategy of taking advantage of the break between southern whites and the Democratic Party, which tended to alienate black voters.

Yet lately there have been signs that things could be changing. Black support for Donald Trump in 2016 and 2020 was higher than was seen for McCain in 2008 and Romney in 2012, reaching 12 percent four years ago, and 19 percent among black men. On top of that, some recent polling has suggested that Donald Trump may be making inroads among the black community, especially among black men.

There are other polls that show that this support may be overstated, and is in fact softer than originally thought. However, even a small shift toward the Republicans could have significant implications in key states. In states like Georgia and North Carolina, where Black voters make up a substantial portion of the electorate, turnout and margins within this group could be decisive. If Trump manages to increase his share of the Black vote — even marginally — it could offset gains Harris makes elsewhere.

Hispanic Voters

Hispanic voters are the fastest-growing demographic in the U.S., and their political leanings are far from monolithic. While traditionally leaning Democratic, recent trends suggest a rightward shift, particularly among Hispanic men.

In Arizona, the Latino population has more than doubled since 2000, now making up 25 percent of eligible voters. Yet the Republicans — particularly Trump — have been making impressive inroads among Hispanic voters, bolstering their chances in the state.

Similarly, in Nevada, Latino voters have shown signs of moving toward the GOP. Factors contributing to this shift include economic concerns, religious values, and perceptions of the Democratic Party's priorities.

The key to the calculus here is that Republicans do not need to win Hispanic voters. Like black voters, they simply need to improve their share of the Hispanic vote, as doing so eats into the Democratic constituency.

If Trump does in fact win, one of the major reasons for the victory will be this burgeoning ethnic realignment. If he loses, then the hope of large scale vote shifts will not have materialized.

The Gender Gap

Interestingly, the shifts mentioned above may be driven by more than just cultural shifts, and identity realignments. There is another demographic shift afoot, and it is helping to drive all of the other changes we are seeing.

Men and women simply aren’t seeing eye to eye about politics anymore.

Historically, women have tended to support Democratic candidates more than men, who often lean Republican. Increasingly, though, this general trend has grown more significant and this year the split between men and women has never been higher.

This is particularly evident among younger voters, as young men and young women dramatically split apart from one another culturally, raising significant concerns about their ability to even date one another.

This gender gap can manifest in many ways electorally, but much of it will come down to turnout. If there is an equal split in men and women’s support for the Republican and Democratic Party, respectfully, than it very well may come down to how many men, and how many women actually vote.

In Pennsylvania, as an example, recent polls show women favoring Harris by a margin of 55 percent to 43 percent, while men favor Trump 54 percent to 44 percent. If 2024 turnout mirrors that of 2020, Harris could win Pennsylvania by a larger margin than President Biden did. Similar patterns emerge in Michigan and Wisconsin, where the combination of women's higher turnout and preference for Harris could secure victories. If, however, a higher share of men vote than in 2020, that would clearly advantage Trump.

Interestingly, in states like Georgia and Arizona, the gender dynamics differ. Men favor Trump by larger margins than women favor Harris, and the turnout gap between men and women is smaller. In these states, Trump could maintain or even expand his lead.

Clearly, the interaction between voter preference and turnout will be crucial.

The Early Voting Calculus

Finally, before I get to my predictions, a word on the impact that early voting is having on the election.

Since the pandemic, early voting has surged in popularity as voters seek the certainty of “banking their vote” without having to worry about anything standing in their way on election day. The rise of early voting, including absentee voting, was seen most distinctly in Democratic voters, but in 2024 there has been an obvious surge of Republicans making use of it.

This trend, however, greatly complicates the interpretation and analysis of candidate performance in various states.

I’ll give you a good example of how.

At almost 2:00 AM this morning, Elon Musk tweeted a screenshot of a spreadsheet with “R minus D early vote deltas in swing states,” showing the shift in early voting rates among both parties between 2020 and 2024. He correctly notes a massive shift in GOP performance, before declaring:

Republican turnout on election day is usually much higher than Democrat, so any state currently leaning R in early voting is very likely to side with Trump. Pennsylvania, while still leaning D, has a massive turnout on Election Day. If the same number vote in person for Trump this time vs last, he will win by ~500k votes (landslide).

He is right that Republican turnout on election day has typically been much higher than Democrats. What Musk fails to consider, though, is whether or not that same advantage will be seen this year.

Republicans could be cannibalizing their election day turnout by voting early, in which case that “day of” advantage will be smaller, or perhaps even gone entirely. If that is the case, all the Republicans will have done is move ahead their vote, rather than expanding the electorate.

The key, in my opinion, is the type of voter who has been voting early. If these early voters are low-propensity or new voters, it could indicate increased enthusiasm for Trump, and a larger GOP vote share overall. If, however, these are high-propensity voters who are simply voting differently this year, Harris will have a much higher chance of winning.

In my opinion, this will be one of the big post-election points of analysis to explain what happened.

Prediction Time: The 2024 Presidential Race

Now that all of that is out of the way, let's dive into the battleground states that will determine the outcome of this election. Based on the data, demographics, and trends, here's how I see things playing out.

Let me start with just a general sense of the election and the mood of the country.

Throw the Bums Out

At the outset, this election should favor any Republican candidate who is challenging a candidate of the incumbent party. Over the course of the last two years, incumbent parties have been getting wiped out by challengers in democracies all over the world. This phenomenon has had nothing to do with ideology, and everything to do with dissatisfaction with the people who have been in power.

In continental Europe, this has taken the form of populist right-wing parties dislodging and disrupting left-wing parties. On most of the continent, leftist governments have been in charge, which has cast the right-wing candidates — particularly the populists — as anti-establishment change agents. As a result, June’s European elections saw major gains by the right in almost every country, while major gains were made in Austria and Germany in national elections this September.

In the United Kingdom, the situation was reversed. The Conservatives had been in power in Britain since 2010, a total of fourteen years, and featured five different prime ministers. The final Tory PM, Rishi Sunak, called a snap election this summer and ultimately lost in a massive landslide to Keir Starmer’s Labour Party, the UK’s mainstream left-wing party.

The point is, voters everywhere are dissatisfied with the people and parties in power, and have been rejecting them at every opportunity.

In the United States, the levers of power are currently held by the Democrats, by and large. It is true that the House of Representatives has a nominal Republican majority, but in the minds of most voters, the presidential administration defines who is in charge, and right now that is Joe Biden.

The Candidates

Donald Trump, for his part, is obviously an imperfect candidate. He has gifts and unique advantages — a plainspoken nature, a disregard for conventional wisdom, an authentically non-robotic personality — that make him, in many ways, very appealing to many people. But he is also crass, childish, uncivil, superficial, insulting, and picks fights he doesn’t need to that actually harm him. I’ve always felt that if he could control his worst impulses, he could have been a fantastically popular president. But he can’t, so he has ended up being divisive.

That said, Kamala Harris is a phenomenally weak candidate. The manner of her selection by the Democrats directly undercuts their rhetoric about “saving Democracy,” but her real liability is her lack of substance. She is always guarded, incredibly thin-skinned, seeks to stage manage and control every interaction, and is allergic to risk. She is clearly afraid of taking difficult questions or thinking on her feet, and voters have a real issue with candidates who seem that empty.

Yes, she has raised more than a billion dollars and the Democratic machine is going to maximize every single vote she is capable of winning. But at a fundamental level, the candidate herself is very limiting to them, and in very key constituencies is going to prevent her from breaking through.

So while Trump is polarizing and alienates many people, he is ultimately helped by the fact that Harris is unlikely to be able to capitalize on this. If another Democratic candidate was running against Trump, this race may be different, but given that it is Harris, I do not see Trump as much of a liability, to be honest. In fact, the strengths he does have may be well matched against her weaknesses, while her strengths, in my opinion, don’t help her much against him.

The Realignment

For much of the last decade, I have been arguing that the United States is in the middle of a major political realignment. This subject deserves its own long-form article, but for those who don’t know American political history, every twenty to sixty years this country has experienced a major shift in the makeup of political parties, who is in them, and why.

The first parties were Federalists, who like Alexander Hamilton believed in a powerful, centralized national government, and a modern industrialized economy, and anti-Federalists, which later evolved into the Democratic-Republicans, who believed in a small, decentralized government and an agrarian based economy.

This is known today as the First Party System — in other words, the first orientation of political parties in this country’s history.

There have been many political alignments since then, including the Second Party System, the Third Party System, the Fourth Party System, the Fifth Party System and the Sixth Party System.

As you get closer to today, the definitions get muddier and harder to pin down to a specific year, or even decade. The Fifth Party System, which was the realignment that came as a result of the Great Depression and Franklin Roosevelt’s New Deal coalition has certainly ended, but when exactly did it end? Did it end with Nixon’s southern strategy in the late 1960s and early 1970s, as the solid south began to abandon the Democratic Party and affiliate with the Republicans? Was it the election of Ronald Reagan in 1980, when the modern GOP began a period of dominance and shifting groups? Was it the 1994 Republican Revolution in Congress?

These things are interesting to debate, but what is clear to most scholars and political observers is that the New Deal Democratic politics that dominated the middle part of the 20th Century certainly gave way to a new system, largely built around the Republican Party capturing southern conservatives and combining them with foreign policy hawks, fiscal libertarians, and religious moralists.

To me, that orientation lasted through the 1980s, 1990s, 2000s and early 2010s.

But starting with the Tea Party wave of 2010 and 2011, another shift began to occur, once again centered around the Republican Party. Slowly, due to populist frustration with both parties, war fatigue, and several other issues, it began to change and started to resemble the pre-World War II Republican Party in many ways. Suddenly, the paleoconservative, Buchananite wing of the party was on the ascendency, as popular discontent and anti-establishment sentiment began to fuel political opinion.

As this happened, the appeal of Republican politicians began to be more working class in character, rather than the blue blooded aristocratic party of the Bushes and Romneys. The issues that were talked about drifted away from the Reagan consensus on issues like taxes and defense, and instead moved toward cultural issues like immigration. The simultaneous rise of progressive extremism on many issues surrounding sexuality and race also drove a rise in right-wing populism in response.

Donald Trump, who is often blamed for turning the Republican Party into “something different,” and who has had to contend with Old Guard figures resistant to his rise, was not the instigation of this phenomenon. Indeed, he was the culmination of it.

Given the changing sentiments of the right, a figure like Trump was inevitable. He simply decided to run at exactly the time that Republican voters wanted an anti-establishment populist firebrand to be their champion.

As Trump conquered the GOP and remade it in his image, he has continued to reshape the voting identities of major demographic groups, both into and away from the Republican Party.

Many people stayed exactly where they were, but many others have been shifting as they have recognized the changes taking place around them. War Hawks and neoconservatives have discovered that they are out of place in Trump’s more non-interventionist party. Free traders are shocked and outraged by the populist demands for tariffs and protectionism. Establishment Republicans do not like the brash, combative nature of the New Right. Voters with college degrees — especially post-graduate degrees — have run away screaming, and suddenly the suburbs are shifting from red to blue.

At the same time, working class voters are more likely to be on board, regardless of race. Trump has been attracting a higher share of not only blue collar whites, but also minority voters as well. Men are increasingly turning to the GOP, as are non-college educated individuals. Anti-war voters, who would not have been anywhere near the Republican Party ten or fifteen years ago, are now flocking to the right, especially as the left has embraced foreign interventionism. Free speech advocates, including many former liberals, have abandoned the Democratic party — which has shown a great hostility to speech of late — and gone to the right where there are now allies.

The point being, voters are rearranging themselves, and resorting their loyalties in new ways. This is a fundamental realignment, and it is obliterating old common assumptions about the future of politics, like the “Emerging Democratic Majority,” which was supposed to be a permanent realignment in favor of the Democrats due to old trends among diverse voters, particularly hispanics.

This permanent majority never materialized, because politics is not static, is always changing, and a major political realignment has been taking shape that was not anticipated at the time that prediction was made.

This realignment helps and hurts both parties, but as a general rule is better for the Republicans than it is for the Democrats. Thus, when you combine it with the other trends I’ve mentioned before, it should provide some wind at the back of Trump and the GOP overall.

In the end: I am predicting that Donald Trump will become the 47th president of the United States. Now let’s examine how.

The Electoral Calculus, State by State

The Sunbelt States

Arizona — Trump Wins

Once a Republican stronghold, Arizona has become a true battleground over the years. That said, the fundamentals of Arizona still remain favorable to the GOP.

Despite the demographic shifts, early voting data shows a strong turnout among Republican voters. The final tallies going into election day showed the Republicans enjoying a 954,306 to 758,342 advantage in returned ballots. Expectations are that Republicans will outvote Democrats on election day while also winning late-arriving ballots. For Kamala to win Arizona she would need to dominate independent voters and have a significant crossover vote of Republicans voting blue.

I find that to be highly unlikely.

Nevada — Trump Wins

Nevada is notoriously tricky to predict, with a high percentage of service industry workers and a significant Hispanic population. Early voting data indicates Republicans have an advantage here as well, which is a departure from previous elections.

Like Arizona, it is likely that there will be a strong Republican showing on election day, even despite the fears over same-day cannibalization. The big wild card here will be the mail ballots from Clark County, specifically how many of them there are and what the overall share of votes is between Trump and Harris. Predictions suggest this will not be enough to overcome what we’ve seen here already. Again, Harris would need giant numbers with independents to win.

Georgia — Trump Wins

Georgia has transformed into a battleground due to demographic changes in the Atlanta metro area, including an increase in college-educated voters and minority populations who have moved there within the last decade.

Harris's underperformance among Black voters compared to Biden in 2020 is a concern. While suburban areas are shifting Democratic, it may not be enough to overcome Trump's advantage in rural areas and potential gains among Black men.

In the end, Georgia is still a nominally red state, and I don’t see any reason to believe there will be a repeat of 2020 here.

North Carolina — Trump wins

North Carolina's diverse electorate includes a mix of urban centers, suburbs, and rural areas. It has some characteristics in common with Georgia, in terms of the shifting electorate, though to a lesser degree.

This is still the state, though, that stuck with Trump even when he was at his most unpopular in 2020, and when Trump fatigue was at its highest. Early voting here has been encouraging, and I don’t see Harris as having the same appeal Biden may have had. I would be shocked if Trump didn’t win North Carolina.

The Blue Wall

Wisconsin — Harris Wins (Maybe)

This is, in my opinion, one of the hardest states to pick this year. Wisconsin was a surprise win for Trump in 2016, largely due to Hillary Clinton ignoring the state, and when it was heavily contested in 2020, the state narrowly went to Biden by only 20,682 votes, or 0.63 percent.

Rural areas and small towns, where Trump has strong support among white voters without college degrees, could offset Harris's advantages in urban and suburban areas. This will be a classic “urban versus rural” duel, and while I absolutely could see Trump winning it, I think Democratic attention to the state and their powerful, well-funded machine-like turnout game will edge them over the finish line and give Harris a win. Maybe.

But you really can’t ignore the fact that Republicans in Wisconsin really know what they’re doing and have an impressive machine of their own. If there is a state I’m giving to Harris that I think might actually go to Trump, it is Wisconsin. In fact, my heart really wants me to give Wisconsin to Trump. So honestly, I’m not even sure if I believe my own pick here.

Michigan — Harris Wins

I’ll keep this one simple: Michigan is less likely, in my estimation, to go to Trump than Wisconsin is, so if Wisconsin doesn’t fall, then neither will Michigan.

It will likely be close, but I happen to think that Governor Whitmer’s political operation is well organized, well funded, and competent, and Harris will be able to take advantage of it. If, on the other hand, Wisconsin falls, Michigan may actually be in play.

Pennsylvania — Trump Wins

This is it. This is the state that matters more than any other states. It is almost cliché to predict that “it all comes down to Pennsylvania,” but in the end I think it actually does come down to Pennsylvania.

This is another state that is at war with itself, with the urban areas — primarily Pittsburgh and Philadelphia — and the surrounding suburbs doing battle with the rural communities.

I was in western Pennsylvania a short time ago, and I was blown away by the vocally and visably pro-Trump voting base there. If you drive closer to the cities, it is exactly the opposite.

The thing that I think makes the difference here is the blue collar character of the state, which I happen to think is a bigger part of their identity than you seen in places like Wisconsin, even though in many ways they are similar.

I think Harris — an awkward, wine-drinking California lawyer with none of Biden’s working stiff charm — is going to have a particularly difficult time with these voters. I think you may see a slightly disappointing performance by her in her core areas, with explosive pro-Trump turnout in the rurals.

This is nothing more than a “gut” call on my part, as there is no legitimate way of knowing. None of the data can really guide us here.

Final Prediction: Trump Wins

287 electoral votes to 251 electoral votes for Harris

Keep Your Eyes On

There are a few states that are not likely swing states, but do have the possibility of surprising some people tonight. Don’t read too much into any of it, but I would keep a sneaky side-eye on Virginia, Iowa, Minnesota and New Hampshire. These states are close enough, and have shown some signs of being interesting… provided there is some kind of minor polling error and the race ends up in one direction or another by a more comfortable margin.

If I'm Wrong, Here's Why

Predicting elections is inherently uncertain, and several factors could lead to a different outcome than I've outlined. As I said before, I don’t really feel great about predicting things this year, because I just don’t think I have the information that would allow me to comfortably make my predictions.

So if I’m wrong, it is because:

• Underestimated Democratic Turnout: If Democratic voters turn out in greater numbers than anticipated, particularly among young people and minorities, Harris could win key battleground states.

• Polling Errors Favoring Harris: The herding effect in polling may have masked a lead for Harris rather than Trump. If polls have overcorrected for past errors, they might be underestimating Democratic support.

• Suburban Surge: Harris could outperform expectations in suburban areas, particularly among college-educated women, offsetting Trump's rural advantages.

• Economic Concerns Shifted: Voters may ultimately decide that Harris is better equipped to handle the economy, especially if they blame Trump for ongoing economic challenges.

• Black and Hispanic Voter Mobilization: Strong turnout and overwhelming support among Black and Hispanic voters could secure victories in states like Georgia, Arizona, and Nevada.

What to Look for in the Returns Tonight to Understand What's Happening

As results begin to come in, keep an eye on these key indicators to gauge the direction of the election:

• Florida's Early Results: Although not a battleground this cycle, early results from Florida's diverse electorate could signal national trends, particularly among Hispanic voters.

• Suburban Counties in Pennsylvania: Watch the margins in Bucks, Chester, Montgomery, and Delaware counties. Harris needs strong performances here to offset rural areas.

• Black Voter Turnout in Georgia: High turnout in counties like Fulton (Atlanta) and DeKalb could indicate a strong night for Harris.

• Maricopa County in Arizona: This populous county includes Phoenix and its suburbs. If Harris is competitive here, she could flip the state.

• Rural vs. Urban Margins in Wisconsin and Michigan: Compare margins in rural counties to previous elections. If Trump is expanding his lead, it could signal victories in these states.

• Early Vote vs. Election Day Vote: Analyze whether early voting patterns are holding or if there's a significant shift in Election Day turnout that favors one candidate.

That’s my take, ladies and gentlemen. Now watch all my predictions go south.

If you are interested in more, this evening I will be anchoring news coverage of the election at Newsradio WGAN from 7 PM to midnight. I’ll be covering the live returns and will be analyzing the results as they happen. Listen live by going to the website, telling Alexa to tune in to WGAN, or just doing it the old-fashioned way by turning it to 560 AM or 98.5 FM.