top of page
Crest_Logo_23-01.png

The puzzle of East Lindsey - revisited

Perspectives


 

Ellen Kirk, Analyst


Wednesday 8 June 2022


Researchers have found that, in general, there is a link between income deprivation and crime. Where incomes are low because people don’t earn much or are out of work levels of offending tend to be higher.


But East Lindsey, a local government district in Lincolnshire, which includes the seaside town of Skegness (pictured), doesn’t follow that pattern.


It is comparatively poor - the 42nd most income deprived local authority in England - but does not experience a lot of crime, ranking 247 out of 317.


It is a puzzle - that Crest Analyst Ellen Kirk has been trying to solve.


In her blog in May, Ellen explored whether a sense of community could be the factor that is helping to keep crime down in East Lindsey. There was quite a response on Twitter.


Criminology student Mark Brown highlighted local responses to crime in the area:




But another user, who has an interest in crime science, offered a different explanation around the idea of "crime generators" and their impact on the local population:




Carina O’Reilly, a Senior Lecturer at the University of Lincoln, agreed with this hypothesis, saying “rurality” was key:


There is one other possible reason, however, that we thought was worth examining in more detail: age.


Gavin Hales, a Senior Associate Fellow at the Police Foundation, argued that age was “more likely” to be the main factor.



Gavin also tweeted an interesting image of the age structure of East Lindsey’s population to illustrate his point further:



In light of this debate, we asked Ellen to carry out some further analysis to see if she could get closer to solving the East Lindsey puzzle.


Age and other factors…


In order to test suggestions that age might be a reason for East Lindsey’s low crime rate, I used a statistical method called ‘regression analysis’. It’s a way of finding relationships and connections between different factors, also known as ‘variables’. In this case I wanted to see what the relationship was between crime and factors such as income, age and the number of people living in an area relative to its size (population density).


The results of the analysis for local authorities in England are shown in the table below.

The key figure to look out for is the ‘R-Squared’ number. It indicates what proportion of the difference between the crime rankings for two local authorities can be explained by a particular factor or factors combined. In simple terms, the closer the R-Squared number is to 1, the closer the link between crime and that particular factor.


The English Indices of Multiple Deprivation (IMD) calculate crime and income deprivation separately for each area and express this as a 'score', before producing a ranking based on these scores. The scores were used here as they are more suitable for regression analyses than rankings.



In terms of individual factors, income score was more closely linked to crime than any other single variable. But combining income with another factor strengthened that link. According to this analysis, therefore, the factors most correlated with levels of crime are income and average age combined, though income together with population density also scores highly.


What about the outliers?


Analysis of data from the Office for National Statistics and the IMD shows the six areas in England with the weakest link between crime and income - where crime is comparatively low, but income deprivation is high. East Lindsey is one of the outliers; the others are Cornwall; Bolsover and Chesterfield in Derbyshire; Torridge in North Devon; and North Norfolk.


In the table below, income deprivation is represented by ‘actual income rank’. The lower the number, the more deprived it is. Offending levels are shown by ‘actual crime rank’: the lower the number, the more crime there is in the area. Here, the predicted scores produced by the regression analysis have been ranked in order to allow for easier comparison with the actual rankings formulated by the IMD.



We can see that income, alone, does not determine the crime level in each of these six areas, though it does push the ‘predicted’ crime rank closer to the ‘actual’ crime rank. When income is combined with the average age of residents there is a much closer relationship with crime levels. In fact, in North Norfolk it’s an almost 100% accurate predictor of that region’s real crime ranking.


The lesson of East Lindsey


When income and age are combined in East Lindsey, the predicted crime rank is nearer to the actual crime rank - the difference is just 66 points compared with a 191-point difference between predicted crime rank according to income score and actual crime rank. This strongly suggests that age is an important factor in keeping levels of crime low in this region of Lincolnshire.


But there are still other factors at play. Could it be that having older folk in an area - who may have lived there for longer - helps embed a sense of community? We’re getting closer to solving the puzzle, but a number of pieces are still missing.

Comments


bottom of page