Mr. Chairman, distinguished members of the committee:
My name is John Samples. I am Director of the Center for Representative Government at The Cato Institute.
I want to thank you Mr. Chairman for inviting me to testify before the committee about election reform.
Mr. Chairman, the United States of America is the greatest example of what James Madison called “popular rule.” We enjoy a legacy of democratic rights and obligations that remains the envy of the world. Differences notwithstanding, we all agree that the franchise is sacred and should be above mere partisan or individual advantage. At the same time, in the spirit of the Founding Fathers, we seek to improve our political system when necessary and possible. I appreciate the opportunity to testify today about some shortcomings of our current electoral system.
In 1994, Congress passed the National Voter Registration Act (popularly known as the “Motor Voter Act”). Congress succinctly stated the aims of the law:
- to establish procedures that will increase the number of eligible citizens who register to vote in elections for Federal office;
- to make it possible for Federal, State, and local governments to implement this Act in a manner that enhances the participation of eligible citizens as voters in elections for Federal office;
- to protect the integrity of the electoral process;
- to ensure that accurate and current voter registration rolls are maintained.
The National Voter Registration Act has clearly fulfilled one of those purposes. Registration rolls grew by 20 percent from 1994 to 1998. Yet enhanced voter registration was never an end in itself. Many activists and experts believed the United States suffered from declining voter participation and that increasing registration would lead to higher voter turnout. Both of these beliefs have turned out to be wrong.
For many years political scientists saw a steady decline in the electoral turnout of the voting age population and as a percentage of registered voters. Leaders in the discipline also thought that reducing the costs of voting — primarily through easier registration — would arrest this steady decline and fortify American democracy. The National Voter Registration Act thus grew out of the findings of political science.
Political scientists have traditionally measured voting turnout as a percentage of the voting age population. Recently political scientists Samuel Popkin and Michael McDonald have shown that “voting age population” is an inaccurate gauge for measuring turnout. The Census Bureau’s estimate of the voting age population includes several categories of persons ineligible to vote: non‐citizens, disenfranchised felons, persons who have moved to a new residence after registration closed, and the mentally incompetent. Popkin and McDonald have produced a new and more accurate measure of the American population eligible to vote. Figure 1 shows Popkin and McDonald’s revised turnout during Presidential elections as percentage of those eligible to vote. Figure 2 indicates revised turnout during off‐year elections as proportion of those eligible to vote.
The United States did see a decline in voting turnout among eligible voters around 1972. Since 1974, the trend in voting turnout in national elections has been basically flat during presidential years and slightly upward during non‐presidential election years. Conventional wisdom to the contrary, the United States had experienced steady turnout at the polls for about three decades. There has been no steady decline, nor a crisis of legitimacy for the American republic. The National Voter Registration Act aimed to solve a problem that did not exist.
“Motor Voter” has also failed to increase voting turnout. Looking at Popkin and McDonald’s revised numbers in Figures 1 and 2, we see that turnout after 1994 is similar to turnout prior to the law. Participation in the Presidential election of 1996 was the lowest since 1948 while estimates of turnout in 2000 suggest an average performance. The same can be said of the off year elections in 1994 and 1998. The world of voting turnout before and after “Motor Voter” looks much the same. This is not really surprising. As the political scientist Martin Wattenberg has pointed out, states like North Dakota, Minnesota and Wisconsin have no or very lenient registration requirements, and yet all have seen declines in voting turnout.
In summary, we have received few of the benefits promised by the National Voter Registration Act. While registration has increased, the law has not enhanced “the participation of eligible citizens as voters in elections for Federal office.” Moreover, the basic premise of “Motor Voter” — reducing the costs of registration would increase voter turnout — now seems disproved by experience. The plus side of the ledger for “Motor Voter” seems empty.
Unfortunately, the same cannot be said of the negative side of the ledger. The costs of “Motor Voter” should be measured by the other goals of the law. Congress intended for “Motor Voter” to both protect the integrity of the electoral process and to ensure officials could maintain “accurate and clean” registration rolls. Neither goal has been met.
The Motor Voter Act allowed citizens to register to vote simultaneously with an application for a driver’s license, by mail, or in person. The Act made it harder to verify the identity of voters seeking to register. It also considerably complicated the states’ task of keeping the registration rolls clean. For example, to remove a voter who has moved from the rolls of a voting district, the local jurisdiction has two choices. First, they could get written confirmation of the move from the citizen. Lacking that, the jurisdiction had to send a notice to the voter. If the notice card was not returned and the person did not vote in two general elections for Federal office after the notice was sent, then the jurisdiction could remove their name from the rolls.
The cost of these mailings is significant. In Indiana, for example, such a mailing would have a price tag of about $2 million or about twice the Election Division’s entire annual budget. Given this price tag and the limited resources of most local election boards, we should not be surprised that the registration rolls throughout the nation are enormously inaccurate. In some counties, election administrators report, the voting roll numbers are bigger than the voting‐age population.
In the short time since Election 2000, we have seen startling new evidence of the disorder of registration rolls in several states. In Indiana, for example, the Indianapolis Star looked closely at the rolls. They concluded that tens of thousands of people appear on the voter rolls more than once, that more than 300 dead people were registered, and that three convicted killers and two convicted child molesters were on the rolls. In general, experts believe one in five names on the rolls in Indiana do not belong there. A recent study in Georgia found more than 15,000 dead people on active voting rolls statewide. Alaska, according to Federal Election Commission, had 502,968 names on its voter rolls in 1998. The census estimates only 437,000 people of voting age were living in the state that year. Similar studies in other states would no doubt return similar data.
In the balance of my testimony, I would like to focus on the events in St. Louis, Missouri, both during the election of 2000 and thereafter. I believe these events point out the real costs “Motor Voter” has imposed on the United States.
Since last fall, “Operation Big Vote” has been active in the St. Louis area as part of a national campaign — promoted by Democrats — to register more African‐American voters and get them to the polling booth. This effort delivered 3,800 voter registration cards to the St. Louis Elections Board on the February 7, 2001, the deadline for the March mayoral primary in that city.
A cursory check of the registration cards turned up questionable names. Shortly thereafter, election board workers spent an entire day calling the names listed on the cards and found that nearly all of them were fraudulent. Many of them sought to register prominent people, dead or alive — as well as at least three deceased aldermen and a dog. The media have reported that close examinations have turned up cards that attempted to register prominent businessmen using their childhood addresses, a former deputy mayor using an old address for an alderman, and a former alderman who has been dead for years. They also found cards for convicted felons and for residents who did not seek to register themselves in the primary. The woman at the center of this vote fraud investigation “doesn’t deny” that some of her canvassers may have turned in bogus voter registration cards. A grand jury convened by St. Louis Circuit Attorney Jennifer Joyce has begun interviewing witnesses regarding the 3,800 bogus registration cards. St. Louis police have obtained a warrant and searched the house of the Operation Big Vote director for evidence.
Not surprisingly, many St. Louis residents are angry that someone had registered them and knew information such as their Social Security numbers. Some of the people registered by the bogus cards told Election Board workers that someone calling himself “Big Mike” came to their homes and said he was with the Election Board and wanted to register them.
This is not the first time Operation Big Vote has been at the center of a voter registration controversy. In 1994, the director of Operation Big Vote was the subject of a similar investigation into fraudulent voter registration cards found among the14,000 that the group had collected to aid a statewide campaign to allow riverboat casinos. No one was indicted.
The implications of the registration fraud scandal in St. Louis are not limited to current events. St. Louis Election Board officials now want to examine 29,500 voter registration cards that came in shortly before the deadline for the November 7 election, in light of discovering that that most of 3,800 cards submitted in February were bogus. John Hancock, executive director of the Missouri Republican Party, called for another look at the last‐minute registrations made for November’s election. He also said he preferred that U.S. Attorney Audrey Fleissig take charge “because I think the federal government can bring more force to bear on an investigation of vote fraud.” The last‐minute registrations last fall could throw into question the close November election in Missouri if a sizable number were fraudulent. Democrat Bob Holden won the Missouri contest for governor by about 21,000 votes statewide. In this way, the loose registration process set up by “Motor Voter” has cast doubt on the integrity and outcome of elections in Missouri last year.
Many will recall the election night controversy in St. Louis. A judge ordered that voting places be kept open late only to be overruled shortly thereafter by a Federal appeals court. The initial ruling accepted the claims of local activists who maintained that thousands of voters had been wrongly placed on an inactive list. As it turned out, local officials had acted properly in composing the inactive list. Missed in the controversy was the fact that up to 400 unqualified voters cast ballots in St. Louis in the 2000 election.
I turn now to the costs paid by the nation as a result of the National Voter Registration Act. I begin with concrete dollars estimates, but I would add that I think perhaps the more important costs have been imposed on the civic culture of the United States.
The clogged rolls have cost taxpayers thousands of dollars in cleanup costs and additional election expenses. For example, the Indiana Election Division has conducted its statewide duplicate program four times at a total cost of about $900,000. Moreover, several county officials in Indiana have increased the number of voting sites unnecessarily because the lists are so inaccurate. The county that includes the transient student population of Indiana University at Bloomington has added about a half dozen precincts since “Motor Voter” became law. Each new precinct costs county taxpayers $10,000 for two voting machines and about $500 per election for additional poll workers and supplies. Statewide in Indiana, more than 200 precincts have been added since the law went into effect, according to state election officials. Such costs are not trivial, especially since the state gets nothing in return for such spending. Such costs for the nation as a whole must be large.
We have also learned about the threat of vote fraud posed by such wildly inaccurate voting rolls. Susan Morandi, Nevada’s deputy secretary of state for elections, noted that the Motor Voter Act made registration easier but also made the process “much more open to voter fraud.” Experts like Deborah Phillips of the Voter Integrity Project add that the trend toward mail‐in and absentee voting exacerbates this problem, since those seeking to manipulate the system can pretend to be a dead person or someone who has moved, and then cast a ballot.
The evidence from St. Louis and elsewhere strongly suggests the reality of registration fraud. Looked at technically, registration fraud is not the same as vote fraud. However, as a practical matter, we should ask why anyone would go to the trouble of committing registration fraud if they did not intend to follow through and commit vote fraud. Otherwise, committing registration fraud becomes a senseless act. Are we to believe that individuals commit registration fraud for thrills or simply as a practical joke? The existence of fraudulent registrations suggests the greater threat of a corrupt election, a danger that we dismiss at our peril. Given the state of the registration rolls, a major vote fraud disaster remains a distinct possibility.
We should not presume that vote fraud is an inconsequential danger. On January 22, 2001, the Miami Herald reported that at least 2,000 illegal votes had been cast in about a third of Florida’s counties — very roughly 6,000 for the state as a whole. On January 9, 2001, it revealed that 452 felons had voted unlawfully in Broward County alone. In Georgia, analysts found that over 5,400 dead people had voted over the past 20 years. As I mentioned, at least 400 unqualified voters cast a ballot in St. Louis last November.
The damage done by vote fraud, of course, is clear. Breaking any law, but especially laws meant to protect the integrity of the electoral process, damages our nation. Vote fraud also devalues the votes of those who register and vote properly. It also strikes many people as unfair: most citizens bear the burdens of exercising the right to vote, those who vote illegally claim the right and wish to escape the minimal burdens associated with that right.
The possibility of vote fraud also harms the nation by calling into question the integrity of our electoral system. The Supreme Court has said that the federal government may regulate campaign finance to prevent corruption or the appearance of corruption. Allow me to suggest that we should similarly be concerned about the appearance of our electoral process. The lax standards for registration encouraged by “Motor Voter” have left the voter rolls in a shambles in many states. As St. Louis shows, the uncertainty surrounding the rolls breeds mistrust and can call the integrity of the system into question. “Motor Voter” has fomented “the appearance of corruption” that has, fairly or not, done real damage to American government. Political scientists have charted the decline in trust in government over the past four decades. I believe “Motor Voter” has been part of that problem, not part of its solution.
The inflation of the registration rolls has also clearly misled Americans about the state of their democracy. Inflated and inaccurate rolls give a false measure of voting turnout as a proportion of registered voters. In fact, we now know that voting turnout as a percentage of registered voters is much higher than we believed because registration rolls are so inflated. In that sense, the news about voting is much better than we thought, and I suspect that we have seen no decline in voting as a percentage of registered voters. We may even have seen a rise in that measure of voter participation.
Finally, politics is about cooperation as well as conflict. The American people expect their representatives to fight hard for a cause but also to make compromises that serve the public good. The general uncertainty surrounding registration procedures — an uncertainty exacerbated by “Motor Voter” — has increased mistrust between the two political parties in Missouri and perhaps elsewhere. Missouri Republicans now believe with plausible reason that some Democrats tried to commit vote fraud last election day. Democrats, on the other hand, suggest Republicans wish to disenfranchise their constituents. Such conflict inevitably weakens our political system and actuates unnecessary partisan rancor that precludes potential bipartisan agreement on some issues.
Mr. Chairman, judged by its purposes, the National Voter Registration Act should be judged a failure. The Act has brought about a substantial increase in the number of registered voters. However, that increase has been bought at a high price. Specifically, the Act has made it difficult if not impossible to maintain clean registration rolls, a major purpose of the law. Moreover, the inaccuracy in the rolls caused by the Act has thrown into doubt the integrity of our electoral system. Finally, the Act has also failed to achieve its other purpose of increasing voter turnout. In sum, the National Voter Registration Act has provided few of its promised benefits and imposed significant costs on the nation. For that reason, “Motor Voter” seems ripe for reform.