It is a real pleasure to read Buddy Roemer. I agree with much of what he says and admire how well he says it. We part ways over policy proposals—but let’s come to that later. His starting points bear attention.

Governor Roemer is neither a laissez-faire libertarian nor a classic campaign-finance reformer. He describes tax policies no one can defend but puts the blame less on special interests than members of Congress. The examples he cites offer an interesting reversal of perspective. For four decades, much of the reform agenda has concerned limits on the role of big money. Governor Roemer opposes this emphasis on principle. Whether one accepts his viewpoint, the Supreme Court has made it pointless in practice to try to silence the rich and powerful. Contributions can still be limited and should be. But trying to make fundamental changes through spending constraints simply won’t pass muster.

Governor Roemer’s approach is to emphasize the role of small donors. Dilute the rich and the powerful by flooding the system with individual contributions, offering candidates an alternative to the way they raise money now. So far I am with Governor Roemer. But how might this be accomplished? This is where we part company.

Governor Roemer says that a presidential candidate who wants to be free from ties to large donors needs only awaken 1 percent of the population to give one hundred dollars each. His math is sound: If 1 percent contributed one hundred dollars each, the result would top $300 million. But while his arithmetic is straightforward, his implementation is not.

Consider that beguiling 1 percent number. According to research published by my colleagues at The Campaign Finance Institute, Barack Obama persuaded about 1.3 percent of the population to give to him before nomination and 1.7 percent over the full two-year election cycle. This was by far the most successful fundraising effort by an individual candidate in history. No one has come remotely close. John McCain was second in 2008 with about one quarter of 1 percent, and Hillary Clinton was third, with less than two tenths of 1 percent. And Obama spent his first nine months raising unprecedented amounts in large contributions to gain the recognition and traction he needed for the small donors to pay attention.

Public matching funds remain the best way to alter candidates’ incentives.

The share of donations becomes increasingly minute for lower offices. The number of donors to a typical House incumbent is well below one quarter of 1 percent of the district’s population. Suppose we quadruple the number of donors and assume one hundred dollars from each. The candidate would net about $700,000, which is about half what a typical winner spent in a House election in 2010. One percent would represent a huge boost in participation, but would not come close to totaling a winning haul. In state elections, CFI’s research shows, most candidates lag further behind the 1 percent target. Were we to add up all of the donors to all of a typical state’s Democratic and Republican general election candidates for governor and for both houses of the state legislatures, we would still end up with less than 1 percent of that state’s total population.

So 1 percent is a tough goal, and it would require major restructuring of their campaigns for most candidates to get there. We cannot expect the candidates to change just because we request it. Something else has to alter their incentives. The Internet is important, but not enough.

Let me suggest three ideas, presented in reverse order of their potential impact. The first two stress the importance of organizations. It is extremely difficult for candidates to maintain and develop a current small-donor list; organizations are better positioned to take on the effort. So, first, we should stop bashing PACs and consider allowing ones that rely solely on small donors to give more to candidates.

Second, let’s free the political parties from their shackles. Under current law parties are allowed to spend as much as they want to help a candidate, but can’t work together with their candidates when they do so. It is ridiculous to force candidates to remain independent of their parties. Parties should be allowed to spend as much as they want in coordination with their candidates, but only from small donations.

Finally the best way to alter candidates’ incentives has proven to be through public matching funds. A recent joint CFI-Brookings-AEI report (“Reform in an Age of Networked Campaigns”) recommended multiple matching funds for small contributions—perhaps a three-to-one or four-to-one match for each dollar of the first one hundred or two hundred dollars that a donor contributes. This would be similar to New York City’s matching fund system, which offers six dollars in public funds for each dollar of the first $175 that a donor gives to each candidate. CFI’s research has shown that the New York City system is highly effective, and it accomplishes a lot of what Governor Roemer wants without violating his stated principles.

Participation in elections is an important public good. Using incentives to encourage participation, therefore, is an appropriate use of public money. It would undoubtedly be a tough sell politically, but so would any program that threatened to alter the status quo. It may be a tall hill to climb, but I agree with Buddy Roemer that the effort is worthwhile.