Become a Member
Jonathan Boyd

ByJonathan Boyd, Jonathan Boyd

Opinion

Hitler’s deadly footprint on post-war Hungary

Sometimes the data are hard to believe, but points to the horrors of the past

July 19, 2018 09:38
Budapest, Hungary - August 14, 2017: Spontaneous memorial for Jewish victims in front of Memorial for victims of the German Occupation in Liberty Square of Budapest.
3 min read

Every so often, data emerging from social surveys stop me in my tracks. I look at a certain number, or proportion, and just know something is awry. It defies logic — it cannot be. There must be a sampling or computational error. So I review it and check it, and check it again. And then, suddenly, the penny drops. There is no error. That bizarre anomaly that inconceivable reality is, in fact, entirely accurate.

I had this experience some time ago when I first started studying Hungarian Jews. And I was reminded of it recently as I was working on a survey for the European Union looking at Jewish people’s perceptions and experiences of antisemitism. Over the past few months, my team has been gathering data in 13 EU member states, including Hungary, and carefully monitoring response levels every 12 hours to continually assess the representativeness of the sample in each country.

One of the ways we have done this is by looking at age distributions the proportions of respondents in each age band because we know, from existing research, what these should look like in most cases.

But the age distributions of the Hungarian Jewish population are known to be odd. There’s one particular age-band that has long appeared to be abnormally small. And sure enough, in this study, we saw it again. About 20 per cent of all respondents were aged in their 40s and another 20 per cent were in their 60s, but only ten per cent were in their 50s. It’s not unusual to see some slight differences in the proportions of different age-bands in a population – indeed, it’s completely normal. But not like that. For one ten-year age band to be half the size of the two ten-year age bands either side of it, there are only three likely possibilities: something has gone wrong with the fieldwork, or with the maths, or something rather unusual has happened to that cohort.