Demos research associate Sean McElwee has a post up at Al Jazeera America, “If everyone voted, progressives would win,” which makes a strong case that recruiting new progressive voters may be a more productive strategy for Dems than targeting likely voters. McElwee explains:

…Progressives do not need a charismatic leader. Instead, they need to invest in unleashing the disgruntled progressive majority. A longer-term strategy for progressives should be to strengthen unions and boost turnout among politically marginalized populations.
“If everybody in this country voted,” the economist John Kenneth Galbraith said, “the Democrats would be in for the next 100 years.” There is strong evidence to support his claim. A 2007 study by Jan Leighley and Jonathan Nagler found that nonvoters are more economically liberal than voters, preferring government health insurance, easier union organizing and more federal spending on schools. Nonvoters preferred Barack Obama to Mitt Romney by 59 percent to 24 percent, while likely voters were split 47 percent for each, according to a 2012 Pew Research Center poll. Nonvoters are far less likely to identify as Republican, and voters tend to be more opposed to redistribution than nonvoters.
In a recent nationwide study, Stockton College professor James Avery found a strong correlation between the electorate’s class bias and the Gini coefficient, a commonly used measure of inequality. In short, the lower the turnout, the higher the class bias and the greater the support for policies that lead to inequality. His study builds on previous research by political scientists Christopher Witko, Nathan Kelly and William Franko showing how class bias in voting reinforces economic inequalities. Their findings are not confined to the U.S. Around the world, voter turnout is correlated with redistributive policies. For example, the turnout of low-income voters has been linked to regressive state tax systems and higher social spending.

While some of McElwee’s conclusions will be familiar to those following voter turnout trends, he includes some interesting new revelations which suggest a new turnout tactic or two. For example, McElwee notes, “72.8 percent of those who do not vote because of weather support the Democratic Party. In fact, weather may have contributed to Electoral College victories for the Democrats in 1960 and the Republicans in 2000.” Improving bad weather contingency plans might make a significant difference in close elections. McElwee cites data which suggest that, in some circumstances, governors may have coattails that help turnout in presidential elections — producing as much as a 6.4 percent edge for Dems.
McElwee also cites overwhelming evidence that “electoral structures dramatically affect turnout.” and that “the more black people in a county — a group that tends to vote for Democrats — the fewer early voting sites there are.” He argues that existing get-out-the-vote campaigns tend to target those who already have high turnout rates, under-investing in turning out new voters who are overwhelmingly inclined to vote for Democrats. And for Dems, election day registration is the pivotal reform, which would increase registration by 18 million nationwide.
Concluding on a hopeful note, McElwee adds “Democrats should mobilize the marginalized progressive majority…Now with Democrats on the defensive across the country, conservatives fighting full franchise and progressives realizing the limits of hero leftism, there may be an effort to mobilize the marginalized progressive majority.”
Perhaps the clincher for McElwee’s contention is that even a small improvement in turnout of these voters could make a huge difference. An extra effort to get marginalized voters to the polls in some key battleground states may be the most cost-effective investment Democrats could make in the 19 months until election day 2016.