PERSPECTIVE ON MEXICO : The New Science of Electoral Engineering : Disputed numbers from the 1990 census set the stage for manipulation in next Sunday’s congressional elections.

Many things are changing in Mexico, but elections and expectations in that regard are not among them. Next Sunday’s vote to elect half of the Senate and all of the lower house, plus governors in six states, is noteworthy for the widespread impression of business as usual that the campaign season has generated.

The results will show a real recovery by the ruling party (the Institutional Revolutionary Party, or PRI), not entirely through democratic practices, but not exclusively as a result of fraudulent, authoritarian practices, either. President Carlos Salinas de Gortari’s party can reasonably expect the middle class’ satisfaction with the stabilization of the economy to translate into support at the polls.

The results also should confirm trends that have emerged in by-elections since 1988: high levels of abstention, as the opposition electorate, both right and left, stays at home, still discouraged by the massive fraud of 1988; and the inability of opposition leaders to unite their followers behind a clear, viable platform.

The most important aspect of this election is that the electorate already knows what it can expect. Mexican voters have seen the future, and it is not a pleasant sight. Under President Salinas, traditional electoral fraud--or “alchemy,” as Mexicans have long called it--while not entirely eliminated, is rapidly being replaced by electoral engineering: a modern, efficient, more subtle and difficult to detect form of mass tampering. It is being done essentially through doctored electoral rolls, manipulation of voter-identity cards, bring-out-the-vote techniques and huge transfers of money and human resources from the government to the ruling party. All of these methods are both effective and well-disguised, although their existence is evident both in electoral procedures and end results.

A good example of this occurred in the process of carrying out the 1990 census, after which, in a separate house-to-house survey, citizens could request to be registered on the voter rolls; later, each would receive a voter identity card. Somewhere in this three-part process, several million Mexicans disappeared. The census revealed that the country had “only” 81 million inhabitants, instead of the 85 million that the authorities, the World Bank and virtually every other observer had estimated. The voting-age population came out to be 45 million. Of these, 39 million requested voter ID cards and 36 million received them, according to official figures. Thus, 10 million voting-age Mexicans will not be able to vote next week.

But there are many doubts about the number of voter cards actually delivered. First, they were handed out late; the original date for full delivery was July 1, in order to give all contending political parties time to review the rolls and check them for repetitions, omissions, padding, etc. Instead, the cards were handed out through July 21, considerably reducing the time available for review.

In Monterrey, the capital of Nuevo Leon, where the entire process took place earlier because the gubernatorial election was held in July, the authorities said that 92% of the cards had been delivered. But in a poll conducted by the respected independent daily El Norte, only 60% of those interviewed actually had them. Indeed, any simple sampling of neighbors, friends or colleagues in Mexico indicates that the government’s 92% delivery claim is high by a wide margin.

Nuevo Leon’s experience on July 7 proved that election-day manipulation remains rampant. Although many observers, including myself, called the voting relatively free and fair, the results confounded everyone. According to the final tallies, nearly 950,000 inhabitants of the state voted, during 10 hours, at 2,100 polling booths--an average of one every minute and 15 seconds. Not only was this a practical impossibility; it also flew in the face of what everyone saw--a healthy turnout, but not nearly that heavy.

Finally, there is the issue of public resources being used for partisan purposes. The opposition, particularly the PAN (National Action Party), has blown the whistle on several minor incidents of misuse of public funds by the PRI. This has not stopped the PRI from broadcasting up to 10 minutes of spots on prime-time television, at a cost of up to 2.5 billion pesos a day, more than the public funds that the opposition parties receive in an entire month.

As a result of these finer features of modernized electoral engineering, there is every reason to expect that Sunday’s vote will provide few surprises. The PRI should get the 60% to 65% that it says it expects; the opposition parties should get anywhere between 7% and 20% each. And a very small number of Mexicans will actually vote, certainly not more than 15 million, perhaps as few as 12 million. With 7 million or 8 million votes in its favor, the official party may possibly recover a two-thirds constitutional majority in Congress, thus ruling a nation of more than 80 million inhabitants with only a handful of votes.

One-party rule seems to be back on its feet in Mexico City.