The tightly contested San Diego Mayoral Election has generated millions of dollars in spending, in a very short amount of time. The election has posed several challenges for candidates and consultants who would normally expect a one to two year run-up before an election of this magnitude.
Early in the process, Political Data constructed several likely voter universes incorporating statewide elections, the stand-alone mayoral special election in 2005, and other factors like data on political donors. While the election is still 24-hours away and we won’t know exact voter turnout figures for several weeks, a look at those voters who have already cast ballots provides evidence that the PDI universes are working. With turnout currently at 34% among those with absentee ballots, the most commonly used likely voter universes are performing at anywhere from 55-66%.The partisan and geographic breakdown has also been interesting for the campaigns and outside observers. This PDI infographic has been updated daily as absentee ballot returns were being reported (over 1200 views!). It shows with mapping and a data table how geography, partisanship and ethnicity are playing out in terms of early absentee turnout. We are looking into creating these kinds of infographics for races in the 2014 cycle–so please check it out and send along any suggestions.
Determining Cost-Efficiency
There are still many absentee ballots coming in, and a full cost-efficiency model can be created after all the votes have been cast, but looking just at the cost of mailing, using a hypothetical 70-cent per-piece mailer, one can see the direct financial benefit of using the PDI universes.
As the following table shows, a hypothetical mailing that went to every household with an Absentee voter would have gotten absolute 100% coverage – (congratulations!). But that came at a cost: an effective cost of $2 per piece, almost three-times the base cost – and a waste of money that could go to other things in the campaign.
Alternatively, a campaign that did their hypothetical city-wide mailing to the 13SDP2 universe would have hit 82% of the actual votes cast so far, but done so at an 88-cent savings per piece.
UNIVERSE | Cost Per Piece (@ .70 cent base) | You Got |
MAIL EVERYONE | $2.06 | 100.0% |
13SDP1 | $1.06 | 65.5% |
13SDP2 | $1.18 | 81.7% |
13SDP3 | $1.27 | 86.9% |
13SDP4 | $1.52 | 92.1% |
13SDP5 | $1.58 | 93.9% |
NOT IN UNIVERSE | $9.45 | 6.1% |
Put into other terms, the MAIL EVERYONE strategy got 100% of the voters, but this hypothetical mailing cost $203,000 compared to the 13SDP3 mailing that cost $108,000. Accounting for the hit rate, the MAIL EVERYONE campaign just wasted $96,000 – enough to send two more mailings to the 13SDP1 universe and have some money left over for the election night party!
And this doesn’t just apply to city-wide mailings. Looking at just a couple other examples, a mailer to ALL Latina voters would cost an effective $3 per piece, but to a targeted universe it would only cost $1.20-$1.30. The cost savings from efficiency would exceed $12,000. A mailing to ALL Republicans under 50 would cost $3.30 per piece, but only $1.40 to $1.50 per piece within a targeted universe, with a net savings of more than $15,000 in mailing costs.