Lehoucq and Molina: Stuffing the ballot box

Lehoucq and Molina. 2002. Stuffing the ballot box: Fraud, election reform, and democratization in Costa Rica. Cambridge: Cambridge University Press.

Puzzle

Why would a government pass laws to eliminate its ability to commit electoral fraud?

Case

Costa Rica over a 50 year period, during which their were four major efforts to reform electoral laws.

• Data about accusations of fraud comes from 123 petitions to nullify electoral results, which record over 1000 (alleged) violations of the 50 years. The authors argue that this is a reliable indicator of the extent of fraud each year, since parties were required to present evidence that fraud actually occurred when they submitted "petitions to annul." [Critique: Parties have incentives to file these accusations only against the winning party's fraudulent activities--but we probably have an underestimate of how much party the opposition parties committed.]
• Accusations of fraud are sorted into four categories along two dimensions. First, whether the accusation involves fraudulent acts (e.g. ballot box stuffing) or procedural irregularities (e.g. closing 15 minutes early). Second, the severity of the act: whether it was a minor or major act.

A politician's dilemma

The dilemma

The model seems similar to Geddes's "Politician's Dilemma." Electoral reforms have long-term benefits, but electoral fraud has short-term benefits. This creates a prisoner's dilemma. Any party that gets into power by cheating fears it will lose power by eliminating cheating.

The solution

So when does electoral reform occur? From pages 11-12:

1. As with Geddes's book, reform can occur when two factions are evenly divided in the legislature. Neither has anything to lose relative to the other by passing electoral reform. This is a necessary condition.
2. Presidents can take advantage of this legislative stalemate to push electoral reform. Electoral reform is in the president's interest for two reasons. First, the president needs to assemble a more national coalition than legislators do, and electoral reform might be a common interest. Second, presidents can run for reelction but not consecutively; thus, electoral reform helps them.
3. Putting 1 and 2 together, presidents can use their veto power to act as a pivot between the two stalemated parties in the legislature, allowing them to push reform through.
4. Finally, electoral fraud has on occasion led to fears of civil war. When it has, the legislature has been especially likely to support reforms.

Two theories to explain the occurrence of fraud

Social and ethnic classes

Upper classes dominate society and use fraud to maintain their dominance. Thus, fraud is most frequent in places with higher social and ethnic stratification. Expect fraud in peripheral areas where poor and minority voters live.

• [Can we really test this hypothesis? Even if fraud were equally likely in every precinct, the wealthier party would have more resources to provide monitors and to file legal claims of fraud. Thus, we would expect claims of fraud to be made in places where the wealthier party has a weaker showing, since that is where the wealthier party will concentrate its resources for detecting fraud.]

Political and institutional conditions

Fraud occurs where competition is most intense, in response to political desires to get in office. And the intensity of competition depends on political and institutional factors in each area.

• [Can we really test this hypothesis? Even if fraud were equally likely in every precinct, wouldn't we expect accusations of fraud to be most common in places that were competing more intensely?]

Magnitude of fraud

• In most elections during the period under study, fraud did not exert enough of an influence to change electoral outcomes. Even if we assume that every ballot at an accused polling station was fradulent, there is still not enough of a change in vote shares to change the presidential outcome.
  • [Critique: Do we really know this? Although the petitions to annul might correlate with the level of fraud (and therefore be a reasonable estimate of where/when fraud is more common), we can't expect them to record every instance of fraud. Surely many instances of fraud go unreported.]
• But still, the fact that a couple elections were (possibly) thrown by fraud is a cause for concern.

Chapter-by-chapter

Chapter 1

Since there are not major social cleavages that are politically salient, politics in Costa Rica is all about controlling the executive and doling out pork. Thus, fraud and violence have become characteristic features of elections.

• Procedural fraud was most common during 1901 and 1912.
• Social structure and electoral laws made fraud most common in the periphery (see above).

Chapter 2

Although there were interests in reforming electoral laws from 1910 to 1914 (when the PR majorly controlled the legislature), it ended up passing only direct elections, not other reforms (previously, presidents were selected by an electoral college, not direct elections). Why? Because one of the PR factions did not want to make the next [presidential] election riskier than it already would be. Passing the direct elections reform is an "efficient" (as opposed to "redistributive") reform--it affects all parties' ability to commit fraud equally. Since direct elections was the only proposed reform that was "efficient," only it passed.

Chapter 3

The switch from an electoral college to direct presidential elections reduced the ability of incumbents to influence selection of the next president. Fraud was rarely a major determinant of outcomes, though in 1923 two PR-dominated Provincial Electoral Councils managed to annul a small number of votes, sending two additional PR members to Congress, and giving the PR a narrow majority in Congress. Since no presidential candidate had won an absolute majority, the Congress got to select the next president from among the front-runners--and as a result, the PR candidate won.

• As earlier, most fraud is in peripheral areas.

Chapter 4

In an effort to legitimize his contested rise to the presidency, the winner in 1923 (Jimenez) pressed Congress to pass electoral reforms. He succeeded in passing a secret ballot. Toward the end of his term, he also succeeded in passing a state-printed (Australian) ballot. Both reforms occurred because Jimenez succeeded in manipulating a narrowly divided Congress; he also "went public," using the media to pressure legislators to go along.

Chapter 5

All these electoral reforms led parties and machines to seek new ways of influencing elections. Ironically, the electoral laws caused parties to turn to increasingly hard-handed and blatant tactics: intimidation, violence, false identification cards, etc.

Chapter 6

Although President Picado was widely suspected of gaining the presidency (1944-8) because of his party's fraud, he soon endorsed a far-reaching Electoral Code (1946). He pushed against the main faction of his party (in Congress) and assembled a coalition to pass the reforms. These reforms probably would not have passed if not for the treat of civil war that hung over the country. Fearing civil war more than a loss of political power, reluctant legislators signed on.

The opposition won the next election, and civil war broke out. Oddly, the 1946 reforms may have actually allowed the opposition to fraudulently win, thus causing the war. At the end of the war, much stronger reforms were passed to protect elections.