The Myth of Ohio

by Cliston Brown | Apr 25, 2020 | Election Analysis

For decades, the notion that Ohio is a bellwether state that decides the outcome of this nation’s presidential elections has persisted. Even today, many Democrats cling to the notion that their party cannot win an Electoral College victory without claiming Ohio’s electoral votes.

This theory is bunk.

Ohio has gained its unearned reputation as an election-deciding bellwether largely due to sheer luck and the fact that its polls close at 7:30 Eastern time. As a result, it is perfectly positioned to be the state that appears to put the Democrats over the top in the Electoral College, but that is merely a function of Ohio usually being called sometime between 10 and 11 p.m. Eastern time, when one candidate or the other is within striking distance of hitting the magic 270 electoral vote threshold.

But Ohio has as much to do with electing presidents as crowing roosters have to do with causing the sun to rise.

In truth, Ohio has not played a decisive role in the election of any Democratic president since Woodrow Wilson narrowly won reelection in 1916. Even in that election, it was not the tipping-point state; the tipping-point state that decided the election was California. (A “tipping point state” is the closest state that gives the winning candidate an electoral college majority.)

It is an arithmetical fact that every Democrat who has won the presidency in the last 100 years would have won an Electoral College majority even if he had lost Ohio. Franklin D. Roosevelt and John F. Kennedy did just that, in 1944 and 1960 respectively.

In fact, Ohio has only been the decisive, tipping-point state in two presidential elections since the Civil War, and in both of those elections (1876 and 2004), Ohio tipped the election to a Republican.

An examination of Ohio’s electoral history demonstrates that Ohio is a Republican-leaning state that almost always votes a few points more Republican than the nation as a whole. Since the Republican Party’s founding in 1854, 41 presidential elections have taken place. Ohio has given Republican presidential candidates greater margins than the national electorate at large has in 33 of those 41 elections, slightly more than 80% of the time. In the 24 presidential elections held in the last century, the Republican margin in Ohio has exceeded the Republican margin in the national popular vote 21 times, exactly 87.5% of the time. In short, Ohio leans to the right of the nation in more than four out of every five presidential elections, and its Republican lean has only become more pronounced with time.

In the 41 elections conducted since the GOP was founded in 1854, Republicans have outperformed their national margin in Ohio by an average of 3.07%. In the last century, the average GOP margin in Ohio has grown to 4.185%. While there are some clear outliers in the data that skew the margin a bit, it is nonetheless clear that Ohio does lean to the right of the national average. In short, if a presidential election is dead-even, or the Democratic nominee wins the national popular vote by roughly a point or less, Ohio will go Republican.

Arithmetically, Democrats simply do not need to win Ohio to win the presidency. By the time Ohio falls into the Democratic column, the election has already been won in other states (even if their polls haven’t closed yet, or there hasn’t been enough of a count in those states to call them for the Democrat by the time that Ohio, an early-closing state, has been largely tabulated).

The implications are simple: Ohio should only be contested if Democrats have plenty of money to spend in other, more crucial states. History conclusively demonstrates that Ohio has never been the tipping-point state in a Democratic victory. To the degree that it ever has been a tipping-point state, it has tipped the election to the Republicans both times.

This is not to suggest that Democrats should give up entirely on Ohio. As long as its major cities provide a treasure trove of votes to the Democratic Party, there will always be a chance to win it, even if that chance is less than 50/50. And just as history shows us that Democrats can win without Ohio (and have done so), history also shows us that Republicans must win Ohio to win the election. They’ve never won nationally without winning Ohio.

But in the later stages of a close presidential race in which tough decisions must be made about resources, Democrats should pull the plug on Ohio without a moment’s hesitation and turn their focus to more fertile ground. Because to put it bluntly: If Democrats find themselves counting on Ohio to win an election, they’re going to lose.

200 Days Out: Biden A Slight Favorite; Democrats Poised to Flip Senate Narrowly

by Cliston Brown | Apr 17, 2020 | Election Analysis

As of today, there are exactly 200 days until the November elections, and at this moment, my expectation is that former Vice President Joe Biden, the Democratic nominee, rates a slight favorite over Republican President Donald Trump.

Based in part on polling averages at this time, and in part on the fundamentals of the race, I currently expect Biden to win 334 electoral votes, compared to 204 for Trump. I refuse to classify any state as a toss-up, which in my mind is a cop-out, but if I were going to list any states as “toss-ups,” those states would be Florida and North Carolina, which I currently list as “tilt Biden.” 

From a polling perspective, Biden currently has an average lead of about 5.5 percentage points, which is roughly 3.4 points above and beyond Hillary Clinton’s popular-vote margin over Trump in 2016. Adding 3.4 points to her totals across the board would flip Florida, Pennsylvania, Michigan and Wisconsin narrowly to Biden, and would turn North Carolina and Arizona into pure toss-ups. I am rating the latter two states as “tilt Biden” for several reasons.

• First, I expect slightly fewer third-party defections nationally in 2020 as compared to 2016, which to me indicates a probable bump of at least a few tenths of a percentage point for Biden.
• Second, there are key Senate races in both states:
  • In Arizona, Democrat Mark Kelly appears to be a clear favorite in his race against incumbent Republican appointee Martha McSally. Given the partisan polarization of the country at this time, I find it hard to imagine there will be very many Kelly voters who will also vote for Trump.
  • In North Carolina, Democrat Cal Cunningham, who is challenging Republican incumbent Thom Tillis for a Senate seat, is remarkably well funded.

I am rating Pennsylvania, Michigan and Wisconsin as “lean Biden” because adding 3.4 points to the Democrats in those states, compared to 2016, and also expecting fewer third-party defections, would flip those states by close but relatively comfortable margins. However, Michigan appears to be much more solid to flip back to blue than the other two, and Wisconsin much less so. I nearly put Wisconsin in the “tilt Biden” category, but was convinced to move it to “lean Biden” by the landslide win by the Democratic candidate in last week’s Wisconsin Supreme Court race, despite the obstacles of voting during a pandemic. It appears clear that there is a motivated Democratic electorate in Wisconsin this year. I also considered putting Pennsylvania in the “tilt Biden” category, but moved it to “lean Biden” based on his long history in the state.

Although the margin in Florida was not significantly greater for Trump in 2016 than it was in Pennsylvania, Michigan or Wisconsin, I am rating Florida “tilt Biden” right now, rather than “lean Biden,” for several reasons:

• First, there is a lot of unpredictability about how its numerous Latino communities will vote. For example: will Venezuelan-Floridians see Democrats as the party of socialism, as many Cuban-Floridians have for generations?
• Second, Republicans typically overperform in Florida relative to polling. I typically assume Republicans will do 2 percentage points better than expected in that state.
• Third, as is the case in most southern states, third parties typically do not do as well there as they do in other parts of the country; third-party defections were less of a factor in Florida to begin with, so Democrats should not expect as significant of a boost from a reduction in “spoiler votes” in Florida as they might in the Rust Belt states.
• Fourth, Trump’s approval ratings in Florida have been notably better than in many other swing states. So I have significantly less certainty about Biden being a favorite in Florida than I have about him being a slight but solid favorite in the three key Rust Belt states that I am rating “lean Biden” at this time.

In addition to Florida, North Carolina and Arizona, I also rate the 2^nd District of Nebraska as “tilt Biden.” It is slightly less Democratic-leaning, on the whole, than the three Rust Belt states of Pennsylvania, Michigan and Wisconsin. But it is also the kind of urban/suburban area where Trump is not especially popular, and at this point, I would rate Biden a very slight favorite to win it.

In the “lean Trump” category, I include Ohio, Georgia, Iowa and the 2^nd District of Maine. A 3.4-percentage point increase in the Democratic totals would not budge any of these jurisdictions out of the Trump column. In fact, a case could be made that Ohio, Iowa and Maine’s 2^nd District should be rated “likely Trump.” However, there tends to be a greater tendency toward sharp swings in the Midwest than in other parts of the country, so it is not inconceivable that Biden could compete in either state. The 2^nd District of Maine is not only demographically similar to Iowa and rural Ohio, but also prone to large swings.

Understand that I consider the “lean Trump” states to be much more likely to go as I expect than any of the “lean Biden” states are. I think Trump is in a stronger position in Ohio, Georgia, Iowa or Maine’s 2^nd District than Biden is in Florida, Pennsylvania, Michigan or Wisconsin.

Moving on to the states rated in the likely category, as a rule, I rated any state that voted for Clinton in 2016 as no worse than “likely Biden.” The states Clinton won by three points or less (Minnesota, Nevada, Maine and New Hampshire) all fall into this category. I expect all of them to be off the board unless there is some sort of significant shift toward Trump nationally, which I do not foresee at this time.

I continue to rate Texas as likely Trump. Despite the ongoing exuberance of many Democrats about Texas, it is too soon to expect Texas to flip in 2020. It may get closer than its nine-point Trump margin in 2016, but not close enough to truly be in play at this time. I expect Biden would need to win nationally by double digits to flip Texas, and I don’t consider that outcome likely.

The remaining states I classify as either “safe Biden” or “safe Trump,” and I doubt very many of those classifications would be considered controversial. I will discuss two of those: Virginia and Colorado, which I rate “safe Biden.” These two states have been consistently moving leftward for a generation, and their demographics (highly urban/suburban, highly educated) do not favor Trump or the Republican Party.

Also rated “safe Biden” are the District of Columbia and the following states: California, New York, Illinois, Washington, Massachusetts, New Jersey, Maryland, Connecticut, Oregon, New Mexico, Hawaii, Rhode Island, Delaware and Vermont.

Rated “safe Trump” are Indiana, Tennessee, Missouri, Alabama, South Carolina, Kentucky, Louisiana, Oklahoma, Arkansas, Kansas, Utah, Nebraska, West Virginia, Idaho, Alaska, Montana, North Dakota, South Dakota and Wyoming.

The breakdown, as I see it today, is as follows:

BIDEN 334

• Safe Biden 210
• Likely Biden 22
• Lean Biden 46
• Tilt Biden 56

TRUMP 204

• Safe Trump 125
• Likely Trump 38
• Lean Trump 41
• Tilt Trump 0

Moving on to the Senate races, I will state that as a rule of thumb, I do not consider any state likely to vote for a presidential candidate of one party and a Senate candidate of the other party. The correlation between presidential vote and Senate vote over the four presidential elections conducted in the 21^st century is around 85%, and in the last presidential election, the same party won the presidential and Senate races in each state which had both races on the ballot.

As a result, at this time, I expect the Democrats to make a net gain of three seats in the Senate, flipping four seats from red to blue and losing the Alabama seat currently held by Democratic Senator Doug Jones. If Biden does, in fact, win the presidency, a +3 Democratic pickup will shift the Senate to Democratic control, with the Democratic vice president breaking the resulting 50-50 tie.

Earlier, I mentioned Democrats Mark Kelly of Arizona and Cal Cunningham of North Carolina as being especially strong candidates, and these are two of the seats I currently expect Democrats to flip. I’m currently rating North Carolina as “tilt Democratic” I also expect Democrat Sara Gideon to defeat incumbent Republican Susan Collins in Maine. In addition to Maine being a likely Biden state, which bodes well for Gideon, Collins’s approval ratings have plummeted as her veneer of supposed moderate has worn off. I currently rate the Arizona and Maine races as “lean Democratic.”

The fourth expected Democratic pickup is also the likeliest one, with Republican Senator Cory Gardner expected to lose in Colorado. As I noted earlier, Colorado is persistently moving left, and the demographics in that state bode poorly for the GOP. Gardner only won his seat narrowly in 2004, running against a Democratic incumbent who ran a poor campaign in a disastrous year for Democrats. Colorado’s continuing leftward shift since then has put Gardner in a decidedly bad position, and I rate Colorado as “likely Democratic.”

Very few other Senate races would not be considered safe at this time for one party or the other. On the Democratic side, I’m rating incumbent Democratic Michigan Senator Gary Peters’ race as “likely Democratic.” On the Republican side, I still rate races in Texas, Montana and Georgia involving GOP incumbents as “likely Republican,” despite the decision by popular Democratic Gov. Steve Bullock to run for the Senate seat in Montana. It is hard to see even a popular Democrat knocking off an incumbent Republican Senator (Steve Daines) while Trump is winning at the top of the ballot in Montana by a wide margin.

I rate two GOP-held seats as “lean Republican,” including Iowa and the seat up for a special election in Georgia. These ratings are the same as the presidential ratings in those states and I expect they will continue to correlate. If either Iowa or Georgia appears to be moving blue at the presidential level, I will adjust the Senate ratings in those states as well.

I am not issuing any House ratings at this time, as that landscape will require further study, but I also consider the House outcome to be in less doubt than the outcome of the presidential race or the composition of the Senate. Unless Trump reverses his fortunes in the suburbs in a big way between now and November, the House is all but certain to remain in Democratic hands.

2018 Midterm Forecast: Democrats +43 in U.S. House; No Net Change in Senate

by Cliston Brown | Nov 5, 2018 | Election Analysis

As I sit down to make my final estimates about the 2018 midterm elections, I have no small sense of trepidation. The last time I attempted to forecast an election, I got it badly wrong. My failure to see what, in retrospect, I feel I should have seen coming in 2016 has made me consider whether my own partisan preferences and/or overconfidence in my own acumen blinded me to the fundamentals of that election.

I resolved after that election to go back to basics, to pay less attention to polls and to do what I have typically done best: to focus on the fundamentals and what I know about the socio-political makeup of each Congressional district and each state.

There are two key fundamentals I have been keeping in mind over the two years of the current cycle:

1. Fundamentally, midterms are always a referendum on the president, and typically, when a president is hovering well under 50% approval ratings, his party loses badly.
2. Special elections during a two-year cycle usually give a very solid indication how the next election will end up.

With that in mind, I think it is clear that Democrats will make major gains on Tuesday. I think it is extremely likely that they will win the U.S. House of Representatives, a large number of governorships currently held by Republicans, and a significant number of state legislative chambers. The U.S. Senate has always been an uphill climb for Democrats this cycle, simply because so many Democratic-held seats (26) and so few Republican-held seats (9) are on the ballot, providing minimal pickup opportunities for Democrats and putting Team Blue on defense all over the map.

So the easiest call of this cycle is that Republicans are likely to retain the U.S. Senate. I see Democratic pickups in Nevada and Arizona, and Republican pickups in North Dakota and Missouri. I will say that my confidence about Arizona and Missouri is slim: Arizona, fundamentally, is still a Republican state, and Missouri’s Democratic Senator, Claire McCaskill, has a history of surviving tough races. But I feel extremely confident about the outcomes in Nevada and North Dakota. I expect when the votes are in, the makeup of the Senate will remain 51 Republicans, 47 Democrats, and two independent Senators aligned with the Democrats.

This means that I expect several endangered Democratic Senators will hold on, including Joe Manchin (D-W.Va.); Jon Tester (D-Mont.); Bill Nelson (D-Fla.); Joe Donnelly (D-Ind.); and Bob Menendez (D-N.J.). It looks like Nelson and Donnelly have opened a slight edge at the end. Democrats are sending out alarms about the troubled Menendez, but it is difficult to see how he loses when Democrats are poised to pick up 3-4 of the five Republican House seats in New Jersey (more on that later). I expect Democratic voters in New Jersey to hold their nose and vote for Menendez, if for no other reason than the Garden State’s disdain for Donald Trump.

In governorships, I expect Democrats to have a big night. It is a slam-dunk that they will flip governorships in Illinois, Michigan and New Mexico. After that, it is a question of who has momentum and enthusiasm. All the signs out of the Midwest is that Democrats will do well there Tuesday, so I expect governorships in Iowa, Ohio and Wisconsin to flip from red to blue as well.

Andrew Gillum, despite recent revelations that are causing his camp heartburn, has the momentum in Florida, and I expect him to prevail to become the first Democratic governor of Florida since 1998.

I expect the governorships of Maine and Nevada to flip, with slightly higher confidence about Nevada.

Now, on to the heavier lifts for Democrats. Republicans in Kansas are sounding alarms about their governor’s race, but the last time they were in peril (four years ago) in the gubernatorial and Senate races, they pulled it out by larger margins than expected. Of course, this isn’t 2014, but the presence of a third-party candidate in the governor’s race leads me to believe that the GOP will narrowly hold on. I also expect that South Dakota is too Trumpy a state to elect a Democratic governor for the first time since 1974, and that Oklahoma is too Trumpy to anoint a Democrat as well.

I also expect that Alaska will flip from independent to Republican, narrowly. Despite independent governor Bill Walker’s decision to drop out in favor of the Democrat, former Senator Mark Begich, Walker will still inevitably win a small percentage of the vote, and that should probably be enough to prevent Begich from winning.

In all, it adds up to a net gain of nine governorships for the Democrats, and a net loss of eight governorships for the Republicans, leaving the states split at 25-25. But watch the following states: the aforementioned Alaska, Kansas, Oklahoma and South Dakota, as well as New Hampshire. And if you want a real long-shot sleeper, keep an eye on Idaho, where Democrat Paulette Jordan appears to be an excellent candidate, but in a state that is very deep Republican red.

State legislatures are not given much attention in the horse race coverage, but based on the current trends and historical performance, I expect the following nine legislative chambers to flip from Republican control to Democratic, narrowing the Republican advantage in legislative chambers controlled from 66-33 to 57-42.

• Maine Senate
• New Hampshire House
• New Hampshire Senate
• Connecticut Senate (from tied to Democratic)
• New York Senate
• Wisconsin Senate
• Minnesota House
• Iowa Senate
• Colorado Senate

If these gubernatorial and legislative forecasts are correct, Democrats would go from full control of eight states to full control of 14 states; Republicans would go from full control of 26 states to full control of 21 states; and 15 states would have split control, as compared to 16 currently.

That brings us to the U.S. House of Representatives, which is the hardest body to forecast. It really comes down to whether one believes the Democrats will continue to overperform as they have in special and off-year elections since 2016, and how much Republican gerrymandering will blunt any wave. Working from east to west, here is where I see changes:

• New York: Democrats net 4 seats.
• New Jersey: Democrats net 4 seats.
• Pennsylvania: Democrats net 3 seats.
• Virginia: Democrats net 3 seats.
• North Carolina: Democrats net 1 seat.
• Florida: Democrats net 3 seats.
• Kentucky: Democrats net 1 seat.
• Ohio: Democrats net 1 seat.
• Michigan: Democrats net 4 seats.
• Illinois: Democrats net 2 seats.
• Iowa: Democrats net 2 seats.
• Kansas: Democrats net 2 seats.
• Texas: Democrats net 2 seats.
• New Mexico: Democrats net 1 seat.
• Colorado: Democrats net 1 seat.
• Arizona: Democrats net 1 seat.
• Utah: Democrats net 1 seat.
• Washington: Democrats net 2 seats.
• California: Democrats net 5 seats.

Going on the state-by-state, race-by-race calls, that puts Democrats at +43, giving them a 238-197 majority in the House.

However, keep in mind that this is an EXTREMELY conservative estimate. I have left a lot of potential Democratic pickups in the Republican column. If Tuesday night is really good for the Democrats, the number of gains could surpass 50. That said, one has to keep in mind the impact of Republican gerrymandering and the fact that even a wave leaves some survivors in the party on the wrong end of the wave. On the flip side, there are always some districts that surprise everybody. To wit: Don Young could lose in Alaska and somebody like Elise Stefanik could get taken out in New York. On the whole, I feel pretty comfortable predicting Democrats +43 in the House. A lot would have to go right for them on Tuesday to exceed that number.

My past results in Congressional elections, dating back to 2006, are within 1.5 seats in the Senate and 4.5 seats in the House. Using my own error margins, that could mean anything from a 51-49 Democratic majority to a 53-47 Republican majority in the Senate, and anything from a 233-202 Democratic House majority to 243-192.