That's not really enough information to calculate exact police voting numbers, as To Vima claims to have done, but it is enough to at least look at trends. And the trend is, as Oikonomides puts it, "astounding." In downtown Athens, for example, most polling stations returned Golden Dawn votes about in line with the national average of 7 percent. But the dozen or so stations clustered around the General Police Directorate received a far higher proportion of votes for Golden Dawn: mostly around 20 or 21 percent, three times the national average.
So where did To Vima get their 50 percent number? They say that 20 to 30 percent of voters in these districts are police -- it's not clear how they know this, maybe police register separately? Assuming that the 70 to 80 percent of non-police voters held to the national 7 percent average, assuming that the variation due entirely to police voters, and assuming that police turnout was the same as non-police turnout, then To Vima estimates that between 45 and 59 percent of policemen in these districts voted for Golden Dawn. From this, they extrapolate that half of Greek policemen support the neo-Nazi group. The math is sound, but it rests on so many assumptions that it's hard to take their much-circulated "50 percent" figure seriously.
Even if you do take their assumptions, it's entirely possible that the prevalence of Golden Dawn voters here is just a coincidence; there might be some other bunch of people in the area voting for the neo-Nazi party. And even if these Greek police did vote for Golden Dawn, that doesn't mean they represent all Greek police. After all, Oikonomides notes that the trend is much less apparent around the national riot police headquarters in Kaisariani, for example.
Still, while it's not a clear national trend, the apparent correlation between police-heavy voting stations and Golden Dawn votes in Athens and elsewhere sure looks like what Oikonomides calls "anecdotal evidence" for a connection.
And there's some reason to think that it might not be shocking for Golden Dawn to find more popular support among policemen than regular Greeks. That's not because Greek police have shown fealty to Nazi ideals or anything like that. Rather, Golden Dawn has some links, mostly ideological and rhetorical, to the far-right military regime that ruled Greece from 1967 to 1974. The junta was brutal, but, like any police state, it was good to the police, and it made law and order a priority. That's not to say that police want a return to dictatorship, but as austerity worsens, a number of government workers could lose their jobs. When Golden Dawn leader Nikos Michaloliakos speaks fondly about the military regime that had made such an ally of police, that could appeal to policemen worried about feeding their families, whether they happen to believe in neo-Nazism or not.