The science of thermodynamics may provide a means of assessing Obama's likelihood of re-election.  Political machines are much like heat engines and tend to obey the same laws.

In thermodynamics, all steam and fuel engines are treated as "heat engines," wherein a source of heat (red) is created and the heat is allowed to flow to the colder "sink" (blue) through the engine (yellow).  Some of the flow is converted into useful work (purple), but a fraction of it must of necessity flow from hot to cold as waste or "increased entropy" (gray arrow) [1].

A tyranny is much like a heat engine.  It extracts labor and goods from its people and gives some of it back -- but keeps a fraction (gray arrow) for the ruling class and the supporting military.  If this "graft" is too small, the regime loses support; if it keeps too much, it runs the risk of either a revolution or a collapse of incentive in the work force.  Tyrants usually rely on a combination of nationalism, repression, isolation, and public apathy to keep the engine working.

In a free society, where the machine depends on popular votes to stay in power, the situation is more complicated.  In the classic Tammany model, votes and taxes are coaxed from the public by promises, but, in addition, campaign money and labor are extracted from special interest groups (green) in exchange for the legislation and funds that the government doles out to them.  These favors, in addition to campaign expenses and graft, constitute the waste flow (gray arrows) that does the public no good but keeps the machine going.

And therein lies the problem: the relative sizes of the arrows.  The general public expects proportionate results for its votes and taxes while the major donors and special interest groups (SIGs) expect a much greater return from their donations, by favorable legislation and/or government spending.  If the gray arrows are too small, the machine will lose the votes and support of its donors and SIGs.  But if the gray arrows become too large, public anger may vote the machine out of office.  Therefore, the machine bosses must optimize a delicate and complex equation to try to satisfy both the public and their supporting SIGs.  Usually, as with some heat engines, there just isn't enough output to justify the input.

But politics is more flexible than thermodynamics.  Like a family living beyond its income, politicians learn to put off their bill collectors by a combination of sops, evasions, and lies.  The actively supporting SIGs are hard to pacify; they expect results.  The machine must address its failure to give SIGs what they want by one or more ploys, including:

• "Look what we've done for you already."

• "We need another term to get you what you want."

• "The opposition blocked us; we can't help you until you give us a clear majority."

• "We're the only hope you've got; the other party hates you."

The public is much easier; if times aren't too hard, public ignorance, inertia, tradition, and suggestibility to hype will override all but the most flagrant abuses of power.  Moreover, the machine usually treats the public as a collection of SIGs and panders to each group's self-interest with appropriate ploys [2].

In 2008, the Obama machine (or "Omachine") skillfully used its kit of ploys to extract votes and support from a wide range of SIGs.  Nonetheless, Obama beat McCain by only 7 percent (53 to 46) -- and that was with the help of the spectacular October surprise of the financial collapse.  Let us look over some of the relevant SIGs and try to estimate how they've changed in the past four years:

• Progressives: With ACORN supposedly dismantled, the Omachine will have to depend heavily on organized progressives, such as MoveOn and 99% Spring, to get out the vote.  Despite boasts of enormous memberships, these organizations have had disappointing participation and are currently begging for volunteers.  They may not be able to get out the vote as well as they did in 2008.  Currently, they are concentrating on telephone calls to the swing states.

• Unions: In 2008, labor unions sent their members out to ring door bells for Obama.  They promised to do the same this year, but union members seem to be responding less enthusiastically, perhaps because of the Omachine's failure to relieve the recession and to enact pro-union legislation.  There may be some losses in support from this SIG.

• Federal employees: This may be a classic demonstration that the principal motivation of most voters is self-interest.  A colleague, a long-time civil servant disgusted by the Obama administration's policies, says that he will probably vote for Obama for fear that Romney will block raises in federal employee salaries and/or cut their pensions.  I suspect that the majority of federal employees think the same way -- with the exception of the armed forces, whose votes the Omachine will try to suppress.

• Gays: Obama left this issue alone in 2008 but recently came out of the closet as definitely pro-gay marriage.  As a result, he has attracted strong gay-rights support but may have alienated some centrist voters.  I'd guess that the Omachine has gained in donations and lost in votes.

• Abortion: The pro-abortion bloc was wholeheartedly for Obama in 2008 and is even more so now.  However, in satisfying this SIG, as by his emphatic and repeated endorsement during the second debate, he has caused an abreaction and energized pro-life groups.  The net effect on votes is hard to predict.

• Elderly: The Omachine has toiled to convince elderly voters that it is the savior of Medicare and Social Security.  But I think seniors are beginning to realize that they have been swindled.  Rising insurance rates and death panels' denials of service are already becoming apparent to many of them.  It may be significant that a recent Pew Research poll shows that older voters are more determined to vote this year than in 2008.

• Young: Young voters were overwhelmingly for Obama in 2008.  According to Pew, and even the Washington Post, the number of young registered voters has dropped, and they are less enthusiastically pro-Obama than in 2008.  The Omachine can expect a small but significant loss of their votes.  More important, their reduced numbers may weaken the get-out-the-vote squads.

• Women: In 2008, women voted for Obama by 56% to 43%.  In 2012, the Omachine, deciding that the women's vote was based primarily on abortion and contraception, launched their War on Women campaign.  It seemed to work; before the presidential debates, according to Pew, women still favored Obama, 56% to 37%.  But the first debate changed everything, leaving Obama and Romney tied at 47%.  The second debate featured women's issues, such as their jobs and abortion.  Polls reflecting the effect of this debate on women's votes may well predict the outcome of the election.

• Blacks: Black voters are still overwhelmingly for Obama, but the percentage and enthusiasm has diminished slightly (from 95% to 92%), especially in swing states such as North Carolina.

• Muslims: In 2008, Muslims favored Obama over McCain 89% to 2%.  One might assume that, because of his Middle East policies, Obama would still have has this group in his pocket.  However, recent polls show a 68%-to-7% ratio, with 25% undecided.  Despite very high turnouts, this is a small group (<4%) that may not significantly affect the Omachine.

• Jewish: Four years ago, the Omachine won 78% of the Jewish vote.  By September 2012, it had apparently diminished to 59%, probably because of the combined effect of the recession and Obama's recent anti-Israel policies.

• Latinos: According to a recent Pew poll, Hispanics favor Obama 69% to 21%.  This is a slight increase from the 2008 ratio of 63% to 31%.  It is particularly significant that Catholic Latinos favor Obama by73%, while Protestant-Evangelical Latinos favor him by 50% to 55%.

• Catholics: As with women, this is a "battleground" SIG.  For the past forty years, the Catholic presidential vote has been split down the middle, with usually less than a 10-percent spread.  In 2008, it favored Obama 54% to 45%.  But Obama's concessions to the pro-abortion bloc has aroused the ire of the Catholic hierarchy and inspired them to unprecedented (albeit inept) political activity.  To counteract this, and to offset GOP Overtures, the Omachine has resorted to dirty tricks to dissuade Catholics from voting for Romney.  On the basis of the above data on Latinos, it seems unlikely that the Omachine will lose many Catholic votes.

I have omitted the SIGs wherein I could find no obvious trend.  My personal conclusion is that the Omachine resembles a leaky engine that has lost some of its original efficiency but is still functioning.  With the possible exception of women, Republicans should not expect any major defections.  Victory will probably be awarded to the party with the greater activism.

[1] This theoretical limit of efficiency is a function of the hot and cold temperatures and cannot be increased by any refinement of design.  It implies that any heat engine must increase the "entropy" (i.e., disorderliness) of the universe -- a property shared by political machines.

[2] The division between promises to the public at large (red) and to donors and special interest groups (green) is not sharp.  A plausible dividing line is whether the promises are made publicly or privately, but even this dichotomy can be fuzzy.