Fixing Our Broken Democracy

Fixing Our Broken Democracy

Stephen H. Unger
October 22, 2009

A large number of Americans, perhaps a majority, recklessly refuses to conform to the views of virtually all established political leaders on an important national issue. They take a position scorned by the mass media and by almost all respected pundits. Fortunately, since very few viable candidates in the last election took this position, no harm was done as these odd-balls had no way to express their outlandish views at the polls.

But wait a minute—maybe I got that wrong! If a lot of people hold certain views, shouldn't they have a chance to vote for candidates promising to implement them? Isn't that a basic democratic premise? It happens that the issue I have in mind is single-payer, Canadian style, health care, but it could have been any of several other important issues. We could argue about whether or not single-payer supporters constitute a majority, but there is no doubt that there are a great many of them (us) [1]. What kind of democracy do we have if a proposal supported by so many people is not even "on the table" for consideration by the executive or legislative branches of government?

With respect to the health care issue, the problem is that the insurance and pharmaceutical industries, principal beneficiaries of the current system, wield enormous influence via campaign contributions, lobbyists, and advertisements. Almost every other important issue is also the subject of similar pressure from corporate interests. Big money also wields political power via its control of the media. This is the likely explanation of the fact that so few newspapers or prominent media pundits support single-payer.

Along with the large role that money plays in distorting the democratic process, there are several other damaging factors. These include e-voting systems, the plurality voting system, and the electoral college. How can we fix our broken democracy? Let's start with campaign finance.

Reduce the Role of Money—At Least a Little

A first step toward reducing reliance of candidates on moneyed interests would be to offer every valid candidate sufficient public funds to wage a reasonable campaign. In order to qualify, we might require a candidate to raise some modest amount of money from constituents, each contributing perhaps $5. This idea has been implemented for state offices in Maine, Arizona, and, more recently, in Connecticut [2, 3]. Recipients who choose this option must agree not to spend additional money in the campaign. If an opponent who has not accepted public funding significantly outspends a recipient, then the public grant is increased to close the gap, up to some limit.

Public support might also include a requirement that TV stations provide some free broadcast time for candidates, possibly some in debate format. Giving all congressional candidates the franking privilege during election campaigns would reduce slightly one of the advantages held by incumbents.

Ideally, strict limits should be imposed on how much a candidate can spend. But there are two problems with this. One is that organizations or people other than the candidate may spend money to publish material supporting the candidate during the campaign. The other problem is that the Supreme Court has ruled that the candidates can spend unlimited amounts of their own money on their campaigns. The Court expressed the view that the first amendment free speech and press provisions imply that restricting monetary expenditures would amount to restricting freedom of speech and of the press. Money = speech seems like a strange equation to me, but its currently the law.

Another fundamental problem is that those who control the mass media can bias election results by slanting their news coverage to favor particular parties or individuals. This is neatly stated by, McGill University history professor Gil Troy: "like it or not, rich people also have the freedom to throw their money around just as the masses have the freedom to throw their collective power around." So, in our "democracy", the phrase, "government by the people" really means, "government by the dollar and the people".

The situation may get even worse as the Supreme Court will soon be hearing a case that could lead to the end of laws prohibiting corporations from making campaign contributions directly. [4]

Giving candidates access to public funds will not really level the playing field. But it would be a significant step in reducing the advantage that wealth can convey. Those without the backing of big money would at least have enough resources to reach significant numbers of voters and perhaps have a fighting chance to win.

Even politicians successful in raising campaign funds in the conventional manner would benefit by switching to public funding, since they would not have to spend so much time fund raising and would be relieved of the burden of kowtowing to donors. Many very capable people unwilling to devote a major portion of their time to fund raising would become available as candidates. "Clean election" laws in Arizona and Maine have indeed been successful for more than a decade. At least one governor and a number of state legislators have been elected using publicly supplied campaign funds.

Getting the Count Right

Even if all other election problems were solved, the mechanics of casting and counting votes must be such that we can be confident that the results have not been corrupted by fraud or error. There is indeed a long history of crooked elections typified by those controlled by Boss Tweed or the Daley machine. Bribery, intimidation, stuffed or stolen ballot boxes, multiple voting, votes cast by cemetery residents were among their favorite tools. In modern American elections, where e-voting systems are common, the same techniques can be, and are, used. But e-voting systems are also vulnerable to much more powerful cheating methods, many of which are nearly impossible to prevent or detect [5].

Given the widespread, successful, use of computer technology, e.g., email, EZ-Pass, ATMs, credit cards, it may seem strange to question the use of computers in the election process. On the face of it, this seems to be an obvious, simple, arena for them to operate in. Unfortunately, this is not the case. Unlike most other computer applications (such as those listed above) involving transactions by individuals, there is no feasible way for voters to verify directly that their transactions were properly processed (i.e., that their votes were properly recorded and actually counted). Furthermore, as is the case for all computer applications, there is no feasible way to verify that the hardware and/or software does not contain concealed features allowing corruption.

Whereas most old fashioned election cheating methods are of a "retail" nature (e.g., each vote corrupted by a bribe involves an interaction between the briber and the bribee), wholesale fraud is possible with e-voting systems. Clandestine features that could be installed or activated by one or two people, could affect scores, hundreds, or even tens of thousands of votes.

So what can be done about this? Surprisingly, this is a matter in which the best direction to move is backwards! Elections based on hand-marked paper ballots, counted by ordinary people, work just fine, as long as the process is carried out openly [6]. The key is that every step must take place under the watchful eyes of representatives of competing organizations (ROCOs).

This requirement is sometimes difficult to satisfy. It necessitates some minimal level of political organization. The warning, "Eternal vigilance is the price of liberty", applies regardless of the election mechanics used. It is important to understand that technology cannot be substituted for this vigilance; if Boss Tweed were left in charge of the polls, he would regard an e-voting system, not as an impediment, but rather as a labor saving device. The good news is that ordinary citizens, serving as ROCOs, have no difficulty monitoring hand-counted paper ballot (HCPB) elections. All aspects are simple and can be made transparent. This approach is used in most other industrialized nations and in large numbers of precincts in several states, including New Hampshire and Maine. We rarely read about election fraud in HCPB jurisdictions. This is not the case for e-voting systems of any type, where even computer experts cannot effectively see what is happening under the hood. It must be emphasized that, in election situations, technology cannot be substituted for citizen alertness. E-voting systems do not deter, they actually facilitate, cheating where citizens are not paying attention, and make corruption feasible even in election districts where there are active ROCOs and where most election officials are honest.

Another anomaly is that, whereas improving the security of a system usually increases its cost, the reverse is true when moving from e-voting to HCPB. Because few areas hold, on average, more than one election annually, e-voting machines, unlike computing systems used for other applications, are idle most of the time. So the associated fixed costs, must be amortised over a small number of elections. HCPB elections are labor intensive, but the fixed costs are very low, so that, in most cases, HCPB elections are actually cheaper than e-voting elections. (This should not be considered as an important factor, given the importance of elections.)

Strangely, the higher expenses associated with e-voting are a significant factor in explaining why such systems are now so widely used. E-voting vendors have much to gain via sales and service contracts. They have been generous in sharing this wealth with those who can help them generate more business [7]. On the other hand, nobody gets rich from HCPB.

Solving the Spoiler Problem

In a typical election to choose one of several candidates for some office, the candidate receiving the most votes wins. This is called "plurality voting" (PV). It is just about the simplest possible system, a strong recommendation. But, particularly when there are multiple candidates for an office, many voters may be faced with a serious dilemma.

A typical case was the 2000 presidential election. Many voters, unhappy with Democratic candidate Gore, thought Nader was a much better proponent of their views. Since it was clear that Nader had no chance to win (if for no other reason than a lack of campaign funds—see above), they were in a quandary. Voting for the candidate they thought best (isn't that what you are supposed to do?), as opposed to voting for Gore, who they disliked, would make more likely a victory by Bush, who they considered the worst candidate. Most Nader supporters held their noses and voted for what they regarded as the "lesser evil". But enough Floridians voted for Nader (about 90,000) to be blamed by angry Democrats for Gore's loss. (They don't blame the over 200,000 Florida Democrats who voted for Bush.)

Clearly it would be desirable to have a voting system such that people would not have to agonize as to whether to vote for the candidate they deemed best. There is indeed a simple change in our electoral system that would accomplish this. Rather than restricting voters to choose one candidate, they could be allowed to choose (approve) as many as they wished. The candidate receiving the most approvals is the winner. This method is called "approval voting" (AV) [8].

Under AV, Nader supporters could have approved both Nader and Gore. More generally, a voter preferring a third-party candidate could, without hesitation, vote for that candidate and then, if sufficiently concerned about a possible win by the "worst" candidate, could also vote for the less disliked contender. The tabulation process is similar to that for PV, altho there is slightly more work since some, perhaps many, ballots will require additions to several candidate totals,

A better known alternative method is Instant Runoff Voting (IRV) where voters rank the candidates [9]. This too would have eliminated the 2000 spoiler problem, but it would usually not do so if the third party becomes a contender. It can also produce bizarre results in some cases. By contrast, AV never produces results that seem irrational in terms of the expressed preferences of the voters. IRV also greatly complicates the tabulation problem, requiring all ballots to be processed centrally or that there be complex exchanges of information between the precincts and a central location. Both methods create more opportunities for fraud and error.

An elaboration of AV is a system called score (or range) voting [10], in which each voter assigns a score to each candidate, e.g., an integer from 0 to 4. The winner is the candidate with the highest total score. This allows voters to express their views more precisely, at the cost of a somewhat more complicated tabulation process. Again the system is well behaved. Under all conditions it makes sense to give one's favorite candidate the highest score and the candidate one dislikes the most, the lowest score. In-between candidates can be given scores reflecting judgments about their merits and chances of defeating one's favorite. There may sometimes be hard decisions to make, but the hardness reflects the actual situation rather than the voting mechanism. In some other systems the difficulty is hidden by a lack of precision.

Other Problems

A serious defect embedded in the constitution is the electoral college, which makes possible serious violations of majority rule in presidential elections. Such violations have actually occurred four times, most recently in 2000. Amending the constitution to repair this defect, so that the president would be elected directly by a nation-wide popular vote does not seem feasible. Getting three quarters of the states to agree seems to be too heavy a burden, particularly since small states often (tho by no means always) benefit from the current system. But a very clever workaround, called the National Popular Vote idea, has been found and real progress has been, made toward implementing it. [11, 12]

Gerrymandering, the calculated manipulation of election district boundary lines so as to bias elections in various ways, is another way in which the democratic process has been corrupted. Here too there exists a nice solution, altho there is not presently an active movement to implement it. [13]

There are good arguments in favor of a more basic change in our governmental system, namely the introduction of proportional representation for our legislative bodies. The object of such a change would be to make these bodies better reflect the views of various groups in the electorate. It is, of course a governmental form widely used in democratic governments elsewhere in the world, particularly in Europe. It is, however, more controversial than the other proposals outlined above.

Doing Something About It

In my opinion, the campaign finance problem ought to be the top priority. It is of critical importance, and, altho it can't be truly solved without a revolution, a meaningful partial solution is not only feasible, but a real start has been made toward implementing it. But others may differ, and, since there are no conflicts among the solutions to the different problems, there is no reason not to attack them simultaneously. In fact, those working on the various issues are likely to be able to cooperate in useful ways.


  1. "Single-Payer Poll, Survey, and Initiative Results", Western PA Coalition for Single-Payer Healthcare, 2009
  2. "Clean Elections", Wikipedia, 10/5/09
  3. S. H. Unger, "Money Talks, and Nominates—and Elects", Ends and Means, 4/10/07
  4. David D. Kirkpatrick, "Court Backs Outside Groups' Political Spending", NY Times, September 18, 2009
  5. S. H. Unger, "E-Voting: Big Risks for Small Gains", Ends and Means, February 5, 2007
  6. S. H. Unger, "Forward to the Past: Junk the Machines, Count Votes Manually", Ends and Means, August 5, 2008
  7. John Gideon, "Corporate Control of the Election Process", Open Voting Consortium, May 15, 2005
  8. "Approval voting", Wikipedia
  9. S. H. Unger, "Instant Runoff Voting: Looks Good—But Look Again", Ends and Means, 3/28/07
  10. S. H. Unger, "Range Voting: Packing More Information into a Vote"March 11, 2008
  11. S. H. Unger, "Electoral Kludge", Ends and Means, 10/8/09
  12. "National Popular Vote", National Popular Vote website
  13. S. H. Unger, "Redistricting: A Nasty Political Problem with a Nice Mathematical Solution", Ends and Means, February 12, 2007

Comments can be sent to me at unger(at)cs(dot)columbia(dot)edu

Return to Ends and Means

site stats