Elections in countries that have been strangers to democracy are never pretty, as Afghanistan's recent vote demonstrated and as any election in Iraq is likely to be. But electoral flaws are not always the whole picture.

Consider Kazakhstan, where five million people cast ballots in September for their parliament's lower house. Western media judged the election a failure. CNN reported it "neither equitable nor free." The Economist snidely referred to the "results" in quotation marks. Most quoted the OSCE's finding that the elections were riddled with "serious shortcomings."

Those elections were flawed. But only a small percent of the breaches were the result of deliberate manipulation. We were there, and visited eight urban and rural polling places, talked with roughly a hundred voters, election officials and observers, as well as officials from six parties. The majority of shortcomings resulted from organizational ineptitude.

The worst cases of outright manipulation occurred when local bureaucrats - all appointed from above - pressured voters to support President Nursultan Nazarbayev's Otan Party. This problem will doubtless persist until more local officials are chosen through elections.

Many, probably most, of the other problems were managerial in nature. Often, voter lists were incomplete or inaccurate. The head of the Central Electoral Commission has now proposed that members of the local public join election officials and administrators in compiling the lists.

Electronic voting, used at 10% of polling stations, also caused problems. Even though voters could choose between electronic and paper ballots, difficulties remained, especially among older voters. Another problem was the bar code system designed to prevent duplicate voting. Some qualified voters didn't get their bar codes; others feared that employers might use them to learn how they voted.

Despite the new election law's assurance of equal coverage in the main media, government TV channels devoted more time to the pro-government parties. Protests by some parties that could not afford the high cost of TV spots and billboards mirror similar complaints in the US.

Far more serious was the lack of transparency in the Central Electoral Commission and the political imbalances on some local commissions. Some of this may have been deliberate, but incompetence also played a part.

So how bad was the election? Exit polls suggest that ballot totals fairly closely reflect voter choices. But how to explain charges of "gross violations" by the Ak Zhol Party, the largest opposition group, or the decision by the Minister of Information, an Ak Zhol candidate, to resign in protest? Or the 89 charges of violations filed by just one bloc?

The ex-minister's charge is disingenuous. He himself presided over many violations, a fact noted by the OSCE back in 1999. He resigned only after winning election to parliament. Clearly, Ak Zhol used accusations of impropriety as a weapon against the government parties. Indeed, many opposition groups engaged in dirty tricks.

Nevertheless, the elections represented significant progress. First, they were competitive. The official registration of ten parties, including opposition groups, surpassed any previous election, and the OSCE admitted that voters had "a real choice." Had systematic manipulation occurred, one-third of the races would not have gone to runoffs.

Moreover, there were real surprises. The Asar Party, founded and led by President Nazarbayev's daughter, was expected to claim second place, assuring her eventual succession. But, despite ample funding and abundant media access, Asar finished third. At the same time, 17% of all candidates were women, far more than in earlier elections. Runoff votes will send more women than ever to Kazakhstan's parliament.

Turnout was relatively high, at 57% of registered, particularly given that the OSCE had intoned that "public interest does not seem to be high"; other experts wrote of "voter fatigue." Defying such sour predictions, voters clearly felt that elections matter. For the first time, parties could make their case in seven debates aired on national television. Many voters changed their choices after watching these.

Another first was that observers from all parties monitored the polling. Thousands of volunteers took their role seriously, producing the complaints that the Prosecutor is now investigating. Another 1,029 foreign observers came from as far away as India and Thailand. The government knew monitors would turn up improprieties, but believes that it has nothing to hide, and much to gain, from honest criticism.

Even the experiment with electronic voting machines was a qualified success. The decision to use them should have been made earlier to allow more training, but most voters experienced no problems. Equally important, no newspapers were closed, stories suppressed, or journalists harassed, as occurred in earlier Kazakh elections.

Obviously, the election's flaws cannot be swept under the carpet. But nor is Kazakhstan seeking to do so. Public debate about how to improve the system has commenced. At the same time, the process as a whole further legitimizes elections in the eyes of people accustomed by Soviet rule to dismissing all votes as charades.

Kazakhstan deserved the criticism it received from Western election monitors. But it also earned kudos that have been withheld. Kazakhstan's elections mark a big step forward. Even an ugly election can be beautiful if it advances democracy - an important lesson for countries like Afghanistan and Iraq.