People pay close attention – maybe too close attention – to political polls for the 2020 presidential election. But you hardly ever hear about how bookmakers handicap the various candidates. The comparison is interesting.

There are two reasons why you might think the betting odds are more accurate than the polls. First, bookmakers only make money if they set their odds in the correct ratio that corresponds to the relative amount of money actually bet on each candidate. Second, people who lay real money down have a lot more at stake than someone answering a poll. Some of them might also have some inside information that the general public doesn’t have access to.

The average numbers of recent national polls listed by RealClear Politics (as of August 26) shows these percentages: Biden 27.2, Sanders 16.7, Warren 16.2, Harris 7.5, Buttigieg 4.8, O’Rourke 2.8, Booker 2.5, Yang 2.0, Gabbard 1.3, Castro 1.5, Klobuchar 1.2, Williamson 1.0. Nothing too surprising there.

But what do the bookmakers say? I found four wagering sites (BetVictor, betWay, Bovada, and NordicBet) that offered political bets and gave a good set of choices. These are all based outside the U.S. The average money lines of candidates for the Democratic nomination who were listed on all four sites were:

  • Elizabeth Warren, +219
  • Joe Biden, +244
  • Kamala Harris, +550
  • Bernie Sanders, +575
  • Pete Buttigieg, +1025
  • Andrew Yang, +1600
  • Tulsi Gabbard, +2625
  • Cory Booker, +3375
  • Hillary Clinton, +3875
  • Beto O’Rourke, +4200
  • Amy Klobuchar, +7000
  • Michelle Obama, +7000
  • Julian Castro, +7250

For those not familiar with money lines, that is the amount you would win if you bet 100 on that candidate. If you divide the number by 100, that is the approximate odds against them winning. For example, the odds are approximately 2.2-to-1 against Warren winning, or, alternatively, she has roughly a 1-in-3.2, or 31%, chance. I say “approximately” since the vigorish (the bookmaker’s edge) built into these payouts mean they cannot be exactly interpreted as the odds of winning.

Other candidates were not listed on all sites. Only two had lower odds than those above: Michael Bloomberg was +4000 and Andrew Cuomo +5000 on betWay. Oprah Winfrey appeared twice, at +10000 and +15000.

One major takeaway from this is that bookmakers, whose livelihood depends on having the pulse of their bettors, think that Warren, with the lowest money line, is the most likely to win, or at least that bookmakers think the betting public thinks she is most likely to win. Another is that Sanders has a smaller chance than the polls suggest. Also, Harris is much stronger, coming in just ahead of Sanders. Converting the odds to approximate win chances, the bookmakers rate Buttigieg, Yang, and Gabbard higher than the polls, while O’Rourke is even weaker than the polls show. And, of course, there is some action for Hillary Clinton and Michelle Obama. For those who think that Hillary is just waiting in the wings and will be anointed as the 2020 nominee, now is your chance to profit: a $100 bet could net a profit of  $3875.

Also note that the two most recent national polls (Economist/YouGov and Monmouth) have shown a shift among the top three: the averages are Biden 20.5, Sanders 19.5, and Warren 19.0); Sanders and Warren have gained at Biden’s expense.  Sanders remains much stronger in the recent polls than the odds suggest. Perhaps the bettors believe that the fix will be in again against Bernie.

For the general election, the average money line was -120 for the Democratic nominee, -110 for the Republican. A negative money line is how much you have to bet to win 100. If you divide it by 100, it is approximately the odds in favor of winning. A more negative number implies being a bigger favorite, so the generic donkey is currently favored. The fact that both numbers are negative just means the bookmakers are getting their cut. Not all sites gave an Other option for the winning party; among the ones that did, the money line ranged from +2500 to +9000 (sorry Libertarians and Greens!).

One other site not included above, MyBookie, seemed out of line compared to the others. The money lines for the top four Democratic candidates there:

  • Biden, +300
  • Harris, +300
  • Warren, +375
  • Sanders, +765

Here Harris shows a lot more support than the other sites. Also, the Republican party was listed as -200 in the general election, i.e., about a 2-to-1 favorite, with the Democrats at +160 and Other at +5000. I don’t know why these are so out of line with the other sites; one would think that arbitragers would have pounced on these differing odds and brought them more in line with each other. By placing appropriately-sized bets at three different sites (Dems +160 on MyBookie, Pubs +109 on Pinnacle, and Other +2500 on Marathon), one could guarantee an 11% profit. That’s my pro-tip for the day.

The MyBookie site also had some Donald Trump specials:

  • Will impeachment begin by 12/31/2019? -5000 no, +1200 yes.
  • Will Trump resign before the end of his first term? -9000 no, +1200 yes.
  • Will Congress obtain Trump’s tax returns by 12/31/2019? -500 no, +300 yes.

Nothing too surprising there, as none of these are given much chance of happening. Then there’s this one:

  • Will Greenland become the 52nd state by the end of Trump’s second term? -140 yes, +100 no!

That’s a bit of an eye-opener, although MyBookie seems to be considering DC to be the 51st state (unless that bet needs somewhere else, like Puerto Rico, to precede Greenland to statehood). Maybe the Free State Project can change their target; Greenland’s population is about 1/24th that of New Hampshire, and should be easy to take over if it becomes a state.

It is still early, and a lot can happen between now and next year. But people who are willing to put their money where their mouth is (or at least the people who are taking that money and want to make a profit) tend to be careful with their prognostications, and my bet (!) is that the bookmaker odds are a better gauge than the political polls.