Confirming fraud in Mozambique’s presidential elections

The 2004 presidential elections in Mozambique were criticised by the main opposition party and international observers, who made allegations of electoral fraud and misconduct. While the available data is limited, such allegations can be tested by examining whether the evidence from polling stations was consistent with specific complaints.

A paper from the CrisisStates Research Centre, in the UK, examines the issue of electoral fraud in the2004 presidential elections in Mozambique. The researchers begin by exploringallegations of fraud at specific polling stations, thenbroaden their analysis to see if the problem was more widespread.

A multiparty democraticsystem was established in Mozambique in 1990. Since then, the country has hadfive multiparty elections, each of which was won by the Frelimo party. In the2004 presidential elections, Frelimo enjoyed a landslide victory over the mainopposition party, Renamo. International observers, the media and oppositionparties (particularly Renamo) complained of electoral fraud and malpractice.

They identified three kindsof misconduct: ballot box stuffing, improper spoiling of ballot papers andintentional organisational failure that meant some voters were unable to vote.Although not enough to change the outcome of the election, the incidents weresaid to happen often.

Ballot box stuffing involvesextra votes being cast or a higher number of votes being recorded to benefit aparticular party or candidate (in this case Frelimo). Where this happened, onewould expect to see an unusually high turnout rate (the percentage of people onthe electoral register who actually vote). Where intentional organisationalfailures made it hard for voters in Renamo-supporting areas to vote, one would expect to see unusually low turnout rates. Whereballot papers were spoilt to damage Renamo’s chances, one would expect a highnumber of invalid votes.

The researchers foundhundreds of polling stations where the evidence is consistent with theallegations of fraud and misconduct:

• 194 pollingstations had turnouts of 85 percent or higher (140 had turnouts of over 95percent).
• 241 pollingstations had turnouts of less than 4 percent and 396 had turnouts between 4 and8 percent.
• Over 25 percentof the votes were invalid at 93 polling stations and between 15 and 25 percentof votes were invalid at 160 stations.
• Suspiciously highand low turnout rates and invalid vote rates occurred mostly in areas whereallegations of electoral fraud had been made.
• At the nationallevel, irregularities in turnout rates and invalid votes were linked to partysupport (for example, high turnout rates were linked with support for Frelimo).

While such statisticalevidence does not prove that electoral fraud took place, the authors concludethat:

• Ballot boxstuffing (benefiting Frelimo) and improper ballot spoiling and organisationalfailure (harming Renamo) did take place on a large scale, even if overallresults were not affected.
• Fraud was mostprevalent in provinces where Renamo registered complaints.
• The analysisprovides a useful method for assessing allegations of fraud and misconduct.
• It alsohighlights areas of concern for future elections in Mozambique and othercountries.