Denominators
Earlier this week, Wikipedia’s front-page “Did you know…” section this week referenced the year that Vatican City had the world’s highest murder rate. In 1998, a double homicide in the Swiss Guards put the murder rate over 200 per 100,000 inhabitants, more than twice the rate in any of the Caribbean countries that usually head the list and roughly infinity times the usual rate in recent years when it has been a separate country. That’s a real rate, with the right denominator, it’s just that the year to year variability makes rates a relatively unhelpful summary.
Vatican City has a denominator problem for minor crimes; in 1992 there were 392 ‘civil offences’ and 608 ‘penal offences’, working out to a bit less than 100% and a bit more than 100% of the number in resident population. The problem is that the tiny state has many millions of visitors per year and they, rather than the resident population, are both the perpetrators and victims of the minor crimes. The resident population isn’t the right denominator.
A less dramatic version of this problem showed up when looking at rates of violent crime in parts of central Wellington some years back.
Thomas Lumley (@tslumley) is Professor of Biostatistics at the University of Auckland. His research interests include semiparametric models, survey sampling, statistical computing, foundations of statistics, and whatever methodological problems his medical collaborators come up with. He also blogs at Biased and Inefficient See all posts by Thomas Lumley »