Welcome to the new version of European Tribune. It's just a new layout, so everything should work as before - please report bugs here.
Those are precisely the figures I was looking for. Thank you. And with growth rates to boot. Vere nice, as that allows me to test for correlation to population growth rates, which might in turn be a proxy for poverty.

Here goes:

It appears that overall crime numbers are constant or declining. In the face of an estimated total population growth of 6.7 percent over the same period, it would appear that the crime rate would be declining.

We further see that across the provinces, murder rate is uniformly down over the period, rape rates are down or constant, while rates of aggravated robbery have increased. The reader is cautioned, however, that either of the years displayed may be anomalous, so these conclusions should be regarded as tentative.

That being said, they do seem to support Nomad's contention that the increase in crime is related to property crime more than violent crime (to the extent that 'aggravated robbery' is not a violent crime, that is).

Further, we see major differences between the provinces, particularly in the case of aggravated robbery. This may be due to different reporting practices, but it may also reflect real differences. Either way, the reader is encouraged to keep this difference in mind when reviewing total crime statistics for RSA.

The murder rate does not seem correlated to population growth, the rape rate might be if you squint and look at it sideways, but it's not a relationship I'd be comfortable about calling a correlation, but aggravated robbery shows something that either is or very strongly resembles a correlation to population growth rate for both 1997 and 2001.

This is consistent with Nomad's observation that the reported increase in crime is due largely to poverty- and property-related crime, at least if we accept that population growth rate is a suitable proxy for poverty.

All in all, Nomad's contentions seem to be supported by the available numbers.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Mon Nov 5th, 2007 at 07:13:16 AM EST
[ Parent ]
It's generally pleasant to see one's own ideas independently validated by a properly done audit - thank you Jake. Much appreciated.

Minor picky questions: How should I read your last three graphs - by province again? Also, what do the symbols signify - which one is the growth rate?

by Nomad on Mon Nov 5th, 2007 at 09:55:24 AM EST
[ Parent ]
I was running out of time when I wrote the last few paragraphs, and I figured (wrongly, of course) that I'd managed to make the graphs self-explanatory.

Each data point represents a single province. The two data series represent 1997 and 2007 (squares and triangles, respectively). The x-axis is the population growth rate is in percent over the four years 1997-2001 (so divide by four to get the annual growth rate, if you neglect compound growth*), and the crime rates (y axis, measured in crimes pr. inhabitant) are normalised to the 1996 census and the 2007 population estimates, respectively.

I decided to omit carjackings for the sake of clarity: They are a subcategory of aggravated robbery, and they make up a sometimes major and wildly varying fraction of the total in that category, which means that the graph for carjackings would not be independent of the graph for total aggravated robbery.

If the graph for carjackings showed a correlation with population growth rates - which it did - the casual reader could have gotten the impression that it provided an independent verification of the property/poverty-related crime hypothesis. This would be wrong, however, since the two graphs would be not be independent. I was not confident in my ability to make this clear from the graphs, and thus decided to omit carjackings altogether to avoid the possibility of misunderstanding entirely.

BTW, if you want to use the figures in this or a future diary (such as the follow-up diary you were hinting was in the pipeline), you're welcome to do so. Just remember to link to the comment I posted them in.

*I haven't done the numbers, so I don't know whether you can get away with that. If you wish to account for compound growth, add 100 %, take the fourth root and subtract 100 % to get the annual rate.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Mon Nov 5th, 2007 at 11:30:04 AM EST
[ Parent ]


Occasional Series