Electoral rigging as partial enfranchisement

The Economist magazine is probably one of the best Western media in English to get to know about poor countries (which of course does not mean that it is the perfect source of information - The Economist is certainly biased towards free market and democracy in a bit naive way). The latest issue includes a survey of Pakistan. The third article in the survey describes how General Musharraf rigged the 2002 election:

Step 1: "[E]nsure that judges in the provinces are sympathetic, because they appoint the magistrates who conduct the polls in the constituencies." This is done by "[Requiring] senior judges to swear allegiance to [General Musharraf] and sack the ones who won't."

Step 2: Tell the returning officers in areas dominated by the opposition to "put polling stations in inaccessible places and select compliant people to staff them."

I've come to think that electoral rigging abundant in today's developing countries is the 21st century form of partial enfranchisement. In Western Europe during the 19th century (and Japan during the early 20th century as well), the suffrage is gradually expanded. First, only rich men are allowed to vote. Then the income threshold for acquiring the voting right is gradually reduced. Later, all adult men gain the suffrage. Finally, all adult women become entitled to vote.

Presidents and prime ministers in today's poor countries cannot do this form of gradual suffrage expansion either because Western donors dislike it or because the enforcement mechanism to ensure that the income threshold requirement is observed is lacking due to the weakness of state capacity.

But poor countries' presidents do not want all adult citizens to vote if they will otherwise be kicked out. (See Llavador and Oxoby (2005) for the formal modelling of this kind of logic.) Therefore, they manipulate the electoral process in a sophisticated way. The location of polling stations as in the above example of Pakistan is one thing. The registration of voters can also be done arbitrarily. This way, only a fraction of citizens who are presumably the supporters of the ruling party are allowed to vote.

If this is true, there is one implication. If citizens engage in retrospective voting - that is, they decide to vote based on the performance of the incumbent - the government has an incentive to target transfers to a certain group of citizens in order to "buy" their votes. During the 19th century, European governments could do this by favouring rich people or men. Today, poor countries' governments can do this by targeting citizens region-wise rather than their socioeconomic status or gender. Why? Because rigging the voter registration or the location of polling stations can only be done region-wise. By targeting government policies to a certain region and allowing citizens in this region to vote, the government can ensure re-election.

This suggests that there must be a correlation between electoral rigging and regional targeting. Which requires empirical investigation.