If the past 56 presidential elections are any barometer, Barack Obama is headed for defeat or a landslide victory. Currently, polls show Obama with a narrow lead over Mitt Romney. He has held that position for most of the year, but if history is an indicator, Obama will exceed the 52.9% of the popular vote and 7.83% margin over John McCain in 2008, or else Romney will be the 45^th president.

After four years in office, the American people have a tradition of either rewarding the president with a greater level of support or else booting him from office. The only time that presidential elections have slightly deviated from this pattern is when third parties have received considerable support. But even then the trend remains consistent.

The Electoral College results are an indicator for this increase support or lose rule, but electoral votes are easily distorted by each state being a presidential election in itself. That is why in 1824, 1876, 1888 and 2000, the popular vote leader failed to get a majority of electoral votes.

Obama might have a difficult time increasing his 365 electoral vote total. Current polls show him somewhere between 300 and 330 electoral votes if all the undecided states are allocated towards him or Romney. That drop in electoral support might be unusual for an incumbent president but not unheard of. However, if the polls that show Obama and Romney essentially tied or Obama with a narrow 1%-2% lead hold up on Election Day, those results would be unprecedented if Obama wins.

First, a look at the 2008 results. Obama received 52.9% to John McCain’s 45.6%. That is a 7.3% margin. Obama won an electoral vote victory of 365 to 173. The minor parties totaled only about 1.5%.

 

In 2000, George W. Bush lost a plurality of the popular vote while achieving 47.9%. Al Gore received 48.4%. Bush received 271 electoral votes, just enough to win. Ralph Nader had 2.7% of the popular vote. He was a factor only in that he pulled some votes from Al Gore, tipping the electoral vote to Bush. In 2004, Nader was not a factor. Yet Bush increased his vote total to 50.7% and his electoral vote total to 286.

Bush’s reelection may have been close, but he did improve his political support. The argument can be made that he never would have been elected president in 2000 if Nader had not run. With that in mind, Bush’s reelection is more significant and establishes the trend that presidents are either reelected with greater support or else defeated.

In 1992, Bill Clinton had 43% and 370 electoral votes. His popular vote margin over George H.W. Bush was a comfortable 5.6%. However, Ross Perot received nearly 19% of the vote and distorted the results. Whether he hurt Bush or Clinton more is still debatable. The election of 1992 marked the defeat of then President Bush, affirming that a president who does not increase his support is doomed to defeat. Bush’s popular vote declined from 53.7% in 1988 when he comfortably defeated Michael Dukakis to just 37.5% in 1992.

In 1996, Clinton received 49.2% and 379 electoral votes, a significant improvement from four years earlier in the popular vote. Clinton also widened his popular vote lead to 8.5%. Perot remained a factor, pulling 8.4%. Again, the trend is affirmed, but this time Clinton increased his support.

In 1984, Ronald Reagan increased his popular vote lead from 9.7% in 1980 to 18.2%. His electoral vote margin also increased from 489 to 525. In 1980, independent John Anderson played a significant role by grabbing 6.6% of the vote. Although Anderson had been a Republican, he may have hurt Carter worse. Regardless, Anderson’s effect on the election had no influence on the winner. Carter was destined to lose. He lost considerable popular support from his narrow victory in 1976 as his vote dropped from 50.1% to 41%.

In 1968, Richard Nixon eked out a victory over Hubert Humphrey 43.4% to 42.7%. George Wallace received 13.5% of the vote. Nixon won the electoral vote 301 to 191 with Wallace at 46. Even if all of Wallace’s votes were cast for Nixon, his 1968 victory is still dwarfed by his 1972 smash of George McGovern. Nixon has 60.7% to McGovern’s 37.5%. The electoral vote was 520 to 17. Clearly, Nixon improved his posiiton from 1968.

Dwight Eisenhower increased his popular vote against Adlai Stevenson from 1952 to 1956 by 55.2% to 57.4%. His electoral vote also increased from 442 to 457. While only a small improvement, it still confirms that presidents increase their support or are defeated.

In 1936, Franklin Roosevelt also increased his margin of victory from 1932. He defeated Alf Landon with 60.8% and 523 electoral votes in the most lopsided presidential election in modern history. In 1932, Roosevelt defeated the incumbent Herbert Hoover with totals of 58.2% and 444. By defeating an incumbent who had been popular in 1928 and increasing his own vote totals, Roosevelt continued the pattern that an incumbent must increase his support or lose.

It is true that Roosevelt’s support fell during his third and fourth runs for the presidency. However, Roosevelt is the only president to seek a third and fourth term. There is no one else with which to compare those results. After eight years in a high profile office, it would be an aberration for a president’s support not to fall. Regardless of the results from those elections, the comparison for 2012 is for presidents who have been in office for four years, not eight or twelve.

The election of 1916 is the first real challenge to the increase support or lose theory. Woodrow Wilson narrowly won reelection with 49.2% of the popular vote and 274 electoral votes. His challenger, Charles Hughes had 46.1% and 254. A socialist candidate grabbed 3.1% but had little effect on the election as his support is unlikely to have gone to either party. In 1912, Wilson had received 435 electoral votes. Thus, it seems that the 1916 election saw a drop in support for Wilson. This is an excellent example where electoral votes distort popular support.

The 1912 election was a watershed in American history. It was the only time since the Civil War that one of the two major parties placed third in a presidential election. Besides Wilson, former President Theodore Roosevelt and incumbent President Howard Taft were running. Plus, socialist Eugene Debbs was on the ballot. Debbs pulled 6%. This election was such an outlier that Debbs gathered 16.5% of the vote in Nevada, running ahead of the fourth place Taft, the sitting president, who had only 15.9%.

 

The 1912 election created the parties as they are today. The Democrats are the more liberal party and the Republican the more conservative. It was not always that way. From before the Civil War, the Republican Party had been the more liberal party as the force against slavery and the party of industrialization. The Democrats were a conservative force backed by rural America and anti-black and anti-immigrant sentiment.

During Roosevelt’s presidency, he began the park system, broke up big businesses and advocated universal health care. After retiring for four years and seeing the anti-progressive policies of Taft, Roosevelt entered the race as the Progressive Party candidate. Roosevelt split the Republican Party. He received 27.4% of the vote to Taft’s 23.2%. Roosevelt also received 88 electoral votes to Taft’s 8.

It is quite possible that if either Taft or Roosevelt had not run, then the other would have defeated Wilson. Only in the South did Wilson run up totals that the combined votes of Taft and Roosevelt could not exceed. Throughout the East, Midwest and West, Wilson picked up electoral votes by narrow margins over Roosevelt or Taft, often achieving victory with slightly more than one-third of the vote.

In 1916, Wilson’s electoral vote total fell to barely enough to secure victory but his popular vote margin increased from 41.8% to 49.2%. Wilson’s electoral landslide in 1912 was a statistical anomaly. It was nothing short of a freak result that is similar to when presidents lose the national popular vote but win the Electoral College, as in the elections of 1876, 1888 and 2000.

Discounting the Electoral College, Wilson’s popular vote totals did increase significantly from 1912 to 1916. Although his plurality over Roosevelt in 1912 was 14.4% and decreased to just 3.1% in 1916, the combined totals of Roosevelt and Taft in 1912 exceeded Wilson’s vote by 8.7%. Thus, Wilson does appear to have expanded his base of support from 1912 to 1916.

Taft’s resounding defeat in 1912 suggests that his strong victory in 1908 would not have been repeated. In 1908, Taft had 51.6% and 321 of the vores. If the 1912 election was only between Taft and Wilson, would Taft have sqeezed out a victory? If he had, it would probably have broken the rule that president’s increase their support or lose. If he had lost, would Wilson’s victory have been narrower than the close race in 1916? We can never know because 1912 was definitely one of the most divisive and odd elections of American history.

In 1900, William McKinley continued the trend by defeating William Jennings Bryan with slightly greater numbers than in 1896. McKinley increased his popular vote share from 51% to 51.6%. His Electoral College margin also increased from 271 to 292.

The next challenges to this rule are the elections of 1884, 1888 and 1892. Grover Cleveland was the Democratic candidate in all three of those elections. He won two and lost one. In 1884, Cleveland defeated James Blaine by the narrow margins of 48.9% in the popular vote to 48.3%. The electoral vote was also close: 219 to 189.

Cleveland’s loss in 1888 to Benjamin Harrison occurred because he did not expand his support. In the Electoral College, Cleveland lost 233 to 168. Although Cleveland increased his popular vote victory margin from .6% to .8%, 48.6% to 47.8%, his overall percentage dropped from the 48.9% of 1884. The 1888 election is the biggest challenge to the increase support or lose theory because Cleveland did increase his popular vote margin of victory, but he also had a decline in total percentage. It leaves a somewhat inconclusive outcome.

However, the argument can be made Cleveland failed to increase his level of support and went down to defeat. By 1892, Cleveland would defeat Harrison 46% to 43%, 277 to 154 in the Electoral College. The Populist James Weaver had 8.5% and 22 electoral votes, taking much of his support from Cleveland in the West and South. In 1892, Harrison also lost support but much more significantly than Cleveland did in 1888, reaffirming that without increasing support a president is doomed to lose.

Ulysses Grant increased his popular vote totals from 1868 to 1872 from 52.7% to 55.6%. His electoral vote total also increased from 214 to 286. Those results are just as expected from a president who achieved reelection.

Abraham Lincoln won a comfortable reelection in 1864 with 55% of the vote to George McClellan’s 45%; 212 to 21 in the Electoral College. That was a vast improvement from 1860 when Lincoln had 39.7% and 180 out of 303 electoral votes.

However, 1860 was even more of an outlier than 1912. Three Democratic candidates split their party’s vote. If the Democratic Party had stayed united, it is likely that Lincoln never would have been elected. Even with fewer states voting in 1864, Lincoln still expanded his popular vote totals from 1.85 million to 2.2 million. Without a doubt, Lincoln increased his level of support.

The next president who would seek reelection is Martin Van Buren, who won convincingly in 1836 with 50.8% of the vote and 170 electoral votes to William Henry Harrison’s 36.6% and 73. That year the Whig Party was split between Harrison and two other candidates who received 12.5% and 46, respectively. By 1840, the Whigs united. It didn’t matter. Van Buren was unpopular. His vote dropped 46.8% and 60 electoral votes. Harrison was again the Whig candidate and received 52.9% of the popular vote and 234 electoral votes. Van Buren did not expand his base of support so he lost.

After a convincing assortment of elections, along comes the election of 1832 which throws a minor wrench into the notion that a president must increase his support or lose.

Andrew Jackson was reelected in 1832, but he received a smaller percentage of the total vote than 1828. In 1828, Jackson had 55.9%. By 1832, he had 54.8%. Although Jackson increased his electoral vote total from 178 to 219, his popular vote support had not increased. Perhaps this is the only clear-cut aberration from a president either increasing his support or losing in a bid for relection. Yet the 1832 presidential election had a third party, the Anti-Masonic Party, headed by William Wirt. It received 7.8% of the vote and was an anti-Jackson party. As easy as it would be to attribute Jackson’s lose of popular support to a third party, it is doubtful that many of Wirt’s supporters would have backed Jackson.

Nevertheless, there are questions at how accurate the popular vote totals were in those years. It had only been since 1824 that popular votes were widely used to elect the president. Before that, the legislatures selected the Electoral College.

In the elections of 1828 and 1832, suffrage was still limited in many states. For example in 1828, Ohio had 21 electoral votes and 158,000 votes cast in the presidential election. Virginia had 23 electoral votes but 46,000 who voted. The roughly equal number of electoral votes but disparity in votes cast is partially explained by the presence of slaves who were counted as three-fifths a person for Congressional representation but could not vote. Yet it does not fully explain for the discrepancy. In many places, particularly the South where Jackson was strong, suffrage was limited. The right to vote was much more widespread in the North. Moreover, South Carolina never did allow a popular vote for its electors until after the Civil War.

As can be seen from this breakdown of the vote by county in 1832, some areas of the South did not vote for president or record the vote by county. By 1836, the non-voting or reporting areas were reduced considerably. By 1844, only a handful of counties outside of South Carolina did not record their vote for the president. In effect, determining the popular support for Jackson in 1828 or 1832 is incomplete and unreliable.

Although the popular vote was used to allocate electors in many of the grayed out states listed below, the lack of county records only highlights the less than formalized voting procedures for a large section of the country.

 

So what is the lesson learned from this? Without a third party candidate to significantly affect the presidential race in 2012, Obama’s popular vote should exceed 52.9% or his 7.3% margin of victory of 2008, or else he should lose. However, the past is no guarantee of what will happen in 2012. Obama has enough of a cushion that he can lose some support and still be elected, but that would be a political anomaly that has never happened before.

It maybe that Obama’s margin of victory in 2008 was padded by the historic nature of electing an African-American as president. Some voters may have cast votes for him that normally would not have voted for a Democrat with his views. This is the reverse Bradley effect. Yet there were probably people who would have voted for a white Democrat that did not vote for Obama.

The current polls have Obama leading Romney by 1%-3%, roughly 47% to 44%. Minor parties are only polling 1-2% of the vote. For Obama to exceed his margin of victory in 2008, he needs to receive all the undecided votes currently available. If that happened, he would probably exceed his electoral vote from 2008 as well. States like Montana and Missouri might fall into the Obama victory column. At this point, that appears unlikely.

On the other hand, for the expand support or lose rule to continue as it has for nearly 200 hundred years in presidential elections, Romney would need a surge of support. The possibility of that happening also appears unlikely as the electorate has been remarkably stable.

In the 1980 election, Reagan and Carter were dead even in the popular vote with just days to go before the election. However, in the last weekend, Carter’s support collapsed. The comparison for the Obama campaign is ominous.

In 1992, Gallup had nearly 6% of Clinton’s support evaporate in the last days of the campaign, with the bulk going to Perot. The possibility of late shifts against an incumbent or a challenger should not be underestimated.

In 1936 and 1948, presidential polling was far less scientific. In both these instances support for the incumbents Roosevelt and Truman exceeded the last polls by nearly 7% in 1936 and 5% in 1948.

The 2012 presidential election can be drawn to match several parallels from past elections. Obama could be like Carter in 1980 and collapse at the end. He could coast to an unexpected comfortable victory like Roosevelt did in 1936 when the economy was also bad. Then there is the possibility of a completely unparalleled result. Romney might win the popular vote but Obama the electoral vote. Obama continues to perform well in the swing states and hold a comfortable electoral lead. Yet Romney trails Obama by a small margin in the popular vote. A shift of only a few percentage points could give Romney a popular vote win but not enough to achieve a majority in the Electoral College.

If history is a lesson in this election year, it is there maybe a sudden shift of support between 4%-7% as Election Day approaches. Whether that will be for Obama or Romney is anyone’s guess. The current polls that show a narrow Obama victory are probably the least likely scenario. Instead, expect either an Obama blowout in the neighborhood of 54%-55% with close to 400 electoral votes or a narrower Romney victory of 51% and 300 electoral votes.