Vermont Sen. has run a terrific campaign, doing far better than the pundits had anticipated. However, he cannot win the Democratic nomination. The math is clear. Counting super-delegates, leads him by more than two to one in the delegate count. He would have to win the great majority of the remaining pledged delegates to overcome her lead, an impossibility given the Democrats' proportional allocation of delegates. Even with Sanders' narrow win in Michigan last Tuesday, Clinton still padded her delegate lead through her blowout victory in Mississippi.

What Sanders can do is elect presumptive Republican presidential nominee by continuing a protracted and contentious nomination battle. That is the verdict of the "Keys to the White House," a historically based prediction system that I developed in collaboration with Vladimir Keilis-Borok, founder of the International Institute of Earthquake Prediction Theory and Mathematical Geophysics. The system has correctly forecast ahead of time the popular vote outcomes of all eight presidential elections from 1984 to 2012.

The Keys consist of 13 diagnostic questions that are stated as propositions that favor re- election of the party holding the White House. When five or fewer of these propositions are false or turned against the party holding the White House, that party wins another term in office. When six or more are false, the challenging party wins.

Right now, the Democrats, who hold the White House have five keys turned against them. This puts them just one key short of a predicted defeat.

The following five keys count against the incumbent Democrats:

• The party's losses in the 2014 midterm elections have cost it Mandate Key 1.
• President Obama's inability to compete for a third term topples Incumbency Key 3.
• The lack of a historic policy change in the second Obama administration forfeits Policy Change Key 7.
• The failure of the American people so far to recognize either the Paris agreements on climate change or the Iran nuclear treaty as a major foreign policy triumph costs the Democrats Foreign/ Military Success Key 12.
• The absence of a candidate with the charisma of a President Franklin Roosevelt or President Kennedy turns the Incumbent Charisma Key 12 against the Democrats.

The following seven keys count in favor of the incumbent Democrats:

• The decision of billionaire former New York Mayor Michael Bloomberg (I) to eschew a third-party or independent campaign for president wins Third Party Key 4 for the Democrats.
• The fact that the United States is in a slow recovery, not an ongoing recession, secures Short-Term Economy Key 5.
• Relatively strong economic growth in the second Obama term as compared to his previous term and George W. Bush's second term locks in Long-Term Economy Key 6.
• The absence of sustained, violent upheavals like those of the 1960s avoids loss of the Social Unrest Key 8.
• The lack of any significant presidential scandal like the Watergate scandal of the 1970s averts the loss of Scandal Key 9.
• The president has not suffered a major foreign policy or military failure, comparable to Pearl Harbor or losing the Vietnam War, keeping Foreign/Military Failure Key 10 in line.
• The Republican challengers lack the charisma of a President Theodore Roosevelt or President Reagan. Even their most dynamic potential candidate, Donald Trump, appeals only to a small segment of the electorate. Thus, Democrats retain the Challenger Charisma/Hero Key 13.

The upshot of this analysis is that the election will turn on the final 13th key, the Incumbent Party Contest Key 2. If the Democrats secure this key, they will be the predicted winners in November. If not, the challenging Republicans will prevail, regardless of the identity of their nominee.

Key 2 is also the single best predictor of election results. Since 1880, when incumbent party nominee Republican James Garfield lost Key 2 but won the presidency by one-tenth of 1 percent in the popular vote, the party in power has endured a major nomination struggle eight times and lost the general election each time.

In contrast, in only three of 25 (12 percent) contests since 1880 has a relatively uncontested nominee lost the popular vote: Republican Herbert Hoover in 1932, Republican Richard Nixon in 1960 and Republican George H. W. Bush in 1992. In 1888 and 2000, two consensus nominees, Democrat Grover Cleveland and Democrat , won the popular vote but lost in the Electoral College. With these two exceptions since 1880, the popular vote winner invariably prevails in the Electoral College as well.

The decision rule for winning Key 2 is a clear one. In the modern era, the party in power secures this key only if the successful nominee wins two-thirds or more of the delegate vote. Although a continued Sanders campaign cannot gain net him nomination, it can deny Clinton the necessary two-thirds majority and forfeit the election to the Republicans.

Don't be deceived for a moment by early general election matchups. Experience with many elections demonstrates that early polls have no predictive value in forecasting general election outcomes. Prediction is feasible only through a system like the Keys that accurately represents how presidential elections really work.

Sanders himself has said that a Republican presidency would be 100 times worse than a Clinton presidency. Indeed, a Republican victory would destroy the legacy of Obama and leave the Democrats a shattered party. Republicans would control the presidency, the Supreme Court, likely both chambers of Congress, and the majority of state and local governments. Everything that a Sanders voter believes in would be lost.

Thus the logic of Sanders's own argument is that the party must unite behind a consensus nominee. It therefore behooves every Democrat, even those who might prefer a Sanders nomination, to back Hillary Clinton.

Lichtman is distinguished professor of history at American University in Washington.