Curtis Gans of the Committee for the Study of the American Electorate has issued his latest report on voter turnout in the United States. There are those who disagree with Gans on some of the finer points of his methodology and on his analyses of proposed electoral reforms, but there's no question that Gans has made a valuable contribution to our discourse and understanding of voter participation for decades. His latest report is no exception.

Looking at turnout in primary elections, both this year and over time, points to one of the particularly disturbing realities of participation in the United States. Although turnout has generally increased or stayed constant in general elections for president and Congress, it has plunged in elections that have an immense impact on our representation in government and in policy decisions. Gans has fingered the problem of sinking turnout in primary elections for years, and this report  shows how fewer than one in five voting-age Americans participated in a statewide primary this year -- despite a number of highly competitive primaries in the Republican presidential race in its early months and despite several significant primaries for U.S. Senate and other offices that will shape Congress in the years to come.

This is consistent with FairVote's findings on turnout in city elections. Turnout in elections for mayor in major cities today is remarkably low. Last November, San Francisco pundits were concerned that turnout in its mayoral race has been only 42.5% of registered voters.  We took a look at turnout in the last election for mayor in our 22 largest population cities and discovered that San Francisco's turnout in fact led the pack. Having Rahm Emmanuel contesting for an open seat in Chicago still left nearly 60% of registered voters on the sideline, for example. Six cities had had more than 85% of registered voters skip the race - including single digit turnout in San Antonio and El Paso.

Such low turnout is disturbing for another reason: it  is not equal across traditional definitions of class. While turnout disparities based on income and education level are significant in our federal elections, they can be simply shocking in local races. As one specific example, the city of Greenbelt, Maryland in 2011 moved its city elections to November and expanded its number of seats as a means to try to boost turnout and expand opportunities for racial minorities to win on a previously all-white city council. While racial minorities ended up being elected, turnout among registered voters ranged from 20% in some precincts to just 2% another. Our exit survey this summer in a Takoma Park city election found that of the 15% of registered voters who came to the polls, 56% had graduate or professional degrees - as opposed to only 14% of residents of the ward.

As a reformer, I like to look to solutions. There's clearly a broader problem of disengagement, cynicism and lack of awareness that goes beyond any single fix or, in fact, any single set of change to voting rules (short of the unlikely step of compulsory voting for such low profile elections). Voting is a communitarian act, and if people don't feel connected to their community or to groups within it involved in elections, they are unlikely to participate. There are, however, some obvious changes to consider. I'll provide three examples of many possible steps:

• Consolidate election dates:If local elections were generally held on one day within a state and if statewide primaries for congressional offices were hold on one day around the nation (the first Tuesday in June, for example), there would be more publicity and voter awareness.
• Voter education and better preparation:We typically run democracy on the cheap in the United States, but some states pay to mail out guides to voters about what's on their ballot and information about the mechanics of voting. That speaks to the value of acting as if we expect participation, rather than be surprised by it. Doing so means making sure every eligible voter is registered to vote through sensible uses of our databases and perhaps a unique national identifier. It means having a thread of learning about government and elections that runs through K-12 schools so no one leaves our schools unprepared to participate. And once people go to the polls, they should have an orderly and efficient experience.
• Reduce the importance of primaries by expanding voter choice:One of the biggest problems of low-turnout in primaries is that they are often decisive for who wins the general election. Most of our legislative elections take place in districts that lean toward one party - not primarily because of redistricting, but because of the underlying partisan landscape of the country. That means the primary winner of the majority party can walk into office. But even in competitive races, it's rare that anyone but one of the major party nominees will win, with the "lesser of two evils" being a familiar refrain.

We could do much more to accommodate and encourage voter choice in the United States. We should replace plurality voting with majority voting systems like instant runoff voting and should replace winner-take-all legislative elections with "fair voting" forms of proportional representation that provide ongoing choice both between and among parties in the general election. Our new report on congressional elections demonstrates the impact of fair voting in a particularly powerful way, showing how every voter would be likely to have meaningful choices in every election while ending up with balanced representation of their district's left, middle and right - a change we can do by statute and grounded in our own electoral traditions.

Such changes and the broader goal of encouraging participation will take a national commitment to voter participation that has been lacking -- particularly in non-November elections like primaries and city elections. We've had enoujgh attention to the problem to know that it's time to take action.