national polls were showing a race that was effectively a tie, the state-level polls in one key state were telling a different story. In Ohio, Obama consistently led Romney for the entirety of the race in nearly every poll, including those conducted at the high-water mark for Romney after the first debate. Of 85 polls conducted in the state during the 2012 calendar year, just nine ever showed Romney leading. Eight showed ties, and Obama led the rest.3
The incumbent’s lead in the state proved durable, and the polls almost exactly nailed his victory margin. The final RealClearPolitics average showed Obama with a lead of 2.9 percentage points, and he would win the state by three points. The accurate Ohio polling was an exception in 2012, though: Obama led on Election Day in the RealClearPolitics average of national polls by less than a point while winning nationally by four, and the averaged polls of many of the other swing states generally undershot Obama’s final margin by roughly three to four points. There was no way to know before the election with certainty that Ohio’s polls would be correct and so many of the national and other state polls would be off. But in hindsight they were a strong sign Obama would win not just Ohio, but nationally as well. Overall, the winning presidential candidate has carried Ohio in 28 of the last 30 elections, so the Ohio polling suggested an Obama win.
While Obama could have won the White House without Ohio, the president’s victory kept alive a dubious streak for Republicans: they still have never won a presidential election without Ohio, going back to the party’s first presidential election in 1856, when Republican John C. Frémont captured the state but lost the election to Democrat James Buchanan.
Not only has no Republican ever won the White House without carrying Ohio, but the state is typically more Republican than the nation as a whole. In the last 30 elections, the Democratic presidential candidate has outperformed his national average in Ohio only six times. This alone should have suggested that either the national polls showing what was basically a tied race, or the Ohio polls indicating a small lead for Obama, were wrong in 2012: it would be historically out of step for a Democrat to significantly outperform his national performance in the Buckeye State.
So the Ohio polling suggested not just an Obama win, but a national victory of more than three points—but not by much more than three points. Ohio’s two-party presidential vote has not deviated more than three points from the national average in any election since the conclusion of World War II. Thus, the Ohio polling would have suggested an Obama win of more than three points but less than six points. Obama won by four.
The long historical record suggests that, with just a few exceptions, Ohio has long mimicked the national voting, and that it has done so better than any other state. The case for Ohio as the nation’s top bellwether state is therefore threefold:
1. Ohio has the best record of any state in voting for the winning candidate.
2. Ohio’s results most often reflect the national voting average.
3. Ohio has provided the decisive electoral votes to the winning candidate more times than any other state.
WINS AND LOSSES
This book focuses on the presidential elections from 1896 through 2012. Why begin there?
The 1896 election began a 30-cycle span, running through 2012, during which Ohio voted for the winning presidential candidate in all but two elections: 1944 and 1960. By percentage, that’s the best record for any state over those 116 years. While Ohio’s voting consistently came close to the national average prior to 1896, a logical place to begin this study is at the beginning of its long and rarely broken streak of voting for presidential winners.
Beyond that, historical and ideological reasons urge starting in 1896. Political scientists have long regarded 1896 as a seminal, realigning election.4 It ended a post-Reconstruction electoral era, from 1876 to 1892, of extremely close presidential elections that featured not one, but two, Electoral College “misfires,” where the winner of the national popular vote lost the presidency (Democrats Samuel Tilden in 1876 and incumbent President Grover Cleveland in 1888). Ohio Republican William McKinley’s more than four-point margin in 1896 was the biggest victory since Republican Ulysses S. Grant’s 12-point reelection triumph in 1872.
More important, 1896 represented an ideological shift in one of the parties. Democratic nominee William Jennings Bryan pushed for the free coinage of silver, co-opting the agrarian Populist movement that had supported James Weaver’s third-party candidacy in 1892, when Weaver received close to 9 percent of the vote and carried five states. Bryan’s silver stance alienated the Democrats’ business wing. The sitting Democratic president, Cleveland, had little use for Bryan, and many Cleveland Democrats deserted the party (including the president, who supported McKinley).
The Democratic Party changed in 1896, even if the voter coalition that supported Bryan looked a lot like the old Democratic coalition (the party’s base would change over time). The political scientist John Gerring, in his study of party ideology in American presidential politics, characterizes the pre-1896 Democrats as a party aligned with the principles of Founding Father Thomas Jefferson, such as promoting “liberty versus tyranny” and opposing the growth of the state. The post-1896 Democrats are a populist party, he argued, concerned with “the people versus the interests.” The Republican Party’s ideological shift from a nationalist party of promoting “order versus anarchy” to a party defined by “the state versus the individual” came later, in the 1920s, likely as a result of Democratic President Woodrow Wilson’s growing of government during his term (cemented by Democratic President Franklin Roosevelt’s liberalism during his more than a decade in office).5 In effect, Bryan—who was also the Populist Party nominee—brought the populists into the Democratic Party, but the Democrats lost more than they gained and remained in the presidential wilderness for 16 years, until a Republican Party split allowed the Democrats to win the presidency in 1912 (and then hold it against a unified GOP in 1916).
There’s something symbolic, too, about 1896 from an Ohio perspective. The country in 1896 came to where Ohio already was. Since the founding of the Republican Party in the early 1850s—the party first produced a presidential candidate, Frémont, in 1856—the GOP carried Ohio in every election through 1896.
That coincided with a golden age for Ohio in national politics. McKinley’s victory in 1896 marked the sixth time in eight elections that a native Ohioan was elected to the White House—Grant twice (in 1868 and 1872), Rutherford B. Hayes (1876), James A. Garfield (1880), Benjamin Harrison (1888), and McKinley. Four years later, McKinley would make it seven Ohio victories in nine tries, before he, like Garfield before him, fell victim to an assassin’s bullet shortly after the beginning of his second term.
Ohio would vote Republican in every election from 1856 to 1912, and so would the nation, save for just two elections: Cleveland’s victories in 1884 and 1892. Into the 1910s and throughout the rest of the 20th century into the 21st, the nation would swing back and forth between the parties, with Ohio almost always close to the national average. More narrowly, given that McKinley was an Ohioan—as was his political Svengali, Mark Hanna—that’s also a reason to start with 1896 here.
Table 2.1 shows how many times each state voted for the presidential winner over the 30 elections from 1896 through 2012.6 Some of the states did not exist in 1896, so their record of voting with winners begins in the first election in which they participated. For instance, New Mexico became a state in time for the 1912 election, so it voted in 26 elections over this time frame instead of 30, like Ohio.
Table 2.1. Electoral record of the states over the last 30 elections, 1896–2012
The most Democratic region in the first half of this 30-election time frame was the South, which voted almost uniformly Democratic from Reconstruction through the presidency of Franklin Roosevelt. A confluence of factors, including the Democratic Party’s increasing policy liberalism, its post–World War II embrace of civil rights activism on behalf of blacks (who began to vote heavily Democratic during Roosevelt’s term in response