Sep 24 2014

Guest Post: @PraxisReform on Harrow’s Crime Statistics

20140924_praxis_reformA brief analysis by @PraxisReform of the effect of the arrival of Ch.Supt. Simon Ovens as Harrow Police Borough Commander, and recent claims made in the press.

An article claiming that the London Borough of Harrow had the lowest crime rate of all London boroughs had been observed, whilst anecdotal evidence from Harrow residents suggested that this situation seemed far from the case.

The low crime claim was based on statistics produced by Harrow police; however these figures are often criticised, for example the report Caught Redhanded: Why we can’t rely on Police Recorded Crime is critical of the target-setting culture within the police, finding that numerical targets drive incentives to mis-record crime, also finding a “lax compliance with the agreed national standard of victim-focused crime recording”.

More serious concerns about the crime-recording process have been set out in an interim report on crime data integrity issued by Her Majesty’s Inspectorate of Constabulary (HMIC).

In that report, HMIC found weak or absent management and supervision of crime recording, significant under-recording of crime, serious sexual offences not being recorded, and some offenders having been issued with out-of-court disposals when their offending history could not justify it.

Now, whilst demonstrating that the Metropolitan police as a whole may or may not be manipulating crime figures in some way is beyond the scope of this analysis, it should be possible however to use the historic crime figures produced by Harrow’s police to examine changes in the nature of Harrow’s police priorities and subsequent results, with a view to seeing whether or not Ch.Supt. Simon Ovens is living up to the praise levelled on him for this recent “achievement”.

It is also worth noting that we never see any kind of statistical analysis or methodological challenges to Harrow police’s crime reporting, from groups that claim to be monitoring and assessing Harrow’s police.

In this analysis, the police’s own longitudinal data will be used (source). The period examined will be from the twelve months prior to the arrival of Ch.Supt. Ovens (April 2013), and the first fourteen months subsequent to his arrival (this being the most recent period that the Metropolitan police’s website would allow to be downloaded), and should allow for changes and trends to be examined over a substantial period of time.

Some use of the police’s Sanction detection data will also be made, as this is arguably the better statistic, since sanction detections are those where some sort of action has been taken against the offender – rather than just a crime logged – which then may or may not get solved at some indeterminate point in the future. However it’s important to note that sanction detection data will always be a lagging statistic, since the slow speed of the UK Courts may result in a lengthy delay between the subject being charged and sanctioned (or found innocent).

Data for the Harrow area was separated out from amongst aggregate data for all London boroughs, and arranged in a spreadsheet, so that Pearson’s r could be calculated for each type of crime for the period before April 2013, and the period after April 2013.

Note: Pearson’s r is a product-moment correlation coefficient, which measures the linear correlation between two variables, although in this case, we are using the function as a very simple guide to whether the trend in the data is tending to be upwards (+1 being very strongly increasing) or downwards (-1 being very strongly decreasing).

The monthly average in the periods before and after April 2013 was also calculated, and large changes found in either calculation will be commented upon below.



Crime figures here seem to have jumped from an average 12.0/month to 29.9/month (an increase of 17.93/month) between the periods before and after the arrival of the new Borough Commander. The subsequent trend is only slightly downwards (r=-0.13). Sanction detections here are also up for Wounding/GBH but only by an additional 6.72/month.

Common Assault & Other Violence


20140924_common_assaultThis is a big concern to many people. It is also the main anecdotal indicator mentioned at the start of this document, where people have noted an ever larger number of fights breaking out in the Harrow area. Sure enough, pre April 2013 the trend was very slightly downwards (r=- 0.12), but subsequently has risen to new recent highs with a trend of (r=0.58). Sanction detections meanwhile have only increased by an average of 5.48/month between the two periods.

A comparison of the average number of monthly offences is 61.2 pre April 2013 and 70.7 post April 2013.

Other violence (whatever that might mean) has also increased after Ch.Supt. Ovens arrival, although it now seems to have declined back to its original level, however the monthly average pre April 2013 number of offences is 10.3, whilst the monthly average post April 2013 number of offences is 18.8. Sanction detections over the same periods have also increased at the broadly same rate (as you would expect to see in this instance).

Sexual Offences


Rape offences seem to have been declining in the year before the arrival of the new Super (r=-0.16), but disconcertingly that trend seems to have reversed (r=0.22). On the positive side, however, the overall numbers seem relatively low, and the increasing trend would suggest that Harrow Police are probably not trying to talk victims into dropping cases, or failing to record this type of crime (an accusation that has previously been levelled at several police forces).

Other sexual offences have remained about level throughout both periods, so this is further evidence that Rape isn’t being downgraded to a lesser class of crime by Harrow police.



The trend in personal robbery was declining pre-Ovens (r=-0.22) but seems to have slumped with his arrival (r=- 0.70), similarly, business robbery was taking a dive over the same initial period (r=- 0.63) but has stayed roughly level during his tenure (r=-0.04). Changes in sanction detection are broadly in line with this decrease.



This is a strange pattern: either Ch.Supt. Ovens reputation as a thief taker has preceded him, or there is something askance with the police’s figures.

You can see from the graph, above, Burglary in a Dwelling was rocketing with a trend r=0.83 prior to April 2013, then subsequent to Ch.Supt. Ovens arrival, has plummeted from an average of 156.4/month to 121.1/month, and falling fast. Other burglaries meanwhile were falling pre Ovens (r=-0.78) but their decline has slowed (r=-0.28) afterwards. This might be a greater focus on protecting residents from burglary, or it might just be that there is nothing left worth stealing from non-dwellings in Harrow. Interestingly, sanction detections for Burglaries in a Dwelling have only increased from 10.0 per month in the first period to 11.2 per month in the second period.

Theft/Taking of Motor Vehicle


Taking of Motor Vehicles was in a downwards trend prior to the arrival of Borough Commander Ovens (r=-0.47), but the figures seems to have initially jumped slightly with his arrival. However, they are now at lower levels, and the trend is continuing sharply downwards (r=-0.62).

Theft From Motor Vehicle


Theft from Motor Vehicles seems to have been in a steep upwards trend pre-Ovens (r=0.48). However, offences have suddenly fallen from 134.8 cases/month to 101.9 cases/month in the period since the arrival of Ch.Supt. Ovens.

Sanction detections however, have fallen by an average of just 17.27 sanctions/month (however this figure is likely skewed by an exceptionally large number of sanctions in this type offence during August/September 2012). In the current period though these thefts have remained broadly constant, though there is a small upwards trend (r=0.25).

Theft from Shops


Theft from Shops, which was in a downwards trend (r=-0.40) before the arrival of Commander Ovens, has jumped from an average 39.6 cases/month to 49.8 cases/month subsequent to his arrival, and has remained stubbornly high throughout the latter period, though in a very slight downward trend (r=-0.11). It might be speculated that this is a reflection of current economic circumstances across the country, with more and more people turning to food banks, those in extreme circumstances may be seeing theft as a viable way of preventing themselves and their families from going hungry. Sanction detections were up between the two periods from 15.8/month to 23.4/month.

Theft from the Person


Theft from the Person figures were rising prior to the arrival of Commander Ovens (r=0.36), but are now falling (r=-0.20). The monthly average figures are 26.4/month pre April compared to 26.7/month post April, whilst sanction detections were up between the two periods from 0.4/month to 0.7/month.

Other Theft


It’s not exactly clear what the police’s “Other Theft” statistics might represent, although these figures were falling pre-Ovens (r=- 0.53), they seem to have levelled off and stayed constant since his arrival (r=0.04). This represents a fall from an average 130.3/month to 118.0/month, with Sanction detections rising from an average 5.0/month to 7.0/month.

Fraud and Forgery


The police categorise this into two groups “Counted Per Victim” & “Other Fraud & Forgery”. In both these categories, the number offences has slumped to near zero levels, after the arrival of Ch.Supt. Ovens. Although these do not account for a huge number of offences, it would seem highly unlikely that the arrival of a new Borough Commander would cause the cessation of virtually all crime in these categories. We must therefore postulate that he has commanded his rank and file Officers to ignore crimes falling into these categories.

Criminal Damage


Here, we see another class of crimes where the overall trend for all types of crime in the class was downwards (r=-0.39), but which has since reversed with the arrival of the new Super. The most striking subcategory here is that of “Other Criminal Damage” where a downward trend of (r=-0.36) has reversed to an upwards trend of (r=0.32). Sanction detections rose from an average 3.7/month to 5.7/month.

Drugs and Drug Trafficking




Incidences of Possession of Drugs have leapt from an average of 49.8 cases/month pre-Ovens to 76.1 cases/month in the latter period. However, incidents of Drug Trafficking have fallen from an average of 7.1 cases/month pre-Ovens to 5.4 cases/month more recently. In respect of sanction detections, these have risen almost in line with the increase, going from an average of 44.8 sanctions/month to 65.5 sanctions/month. Whilst Drug Trafficking sanctions have remained constant at 5.5 sanctions/month in both periods.

But, this situation does not make sense, because if there are more offences of drug possession, it must follow that there would be more trafficking of drugs, in order to supply those possessing the drugs. Whilst if the increase in offences is purely down to better policing, then you would expect to see a concomitant increase in offences and sanction for Drug Trafficking. Therefore the only possible answer here is that cases of drug trafficking are being downgraded into cases of drug possession, in order to massage the overall figures downwards.


The sudden drop in the absolute number of offences for many crimes (Burglary, Fraud, Theft from Motor Vehicles etc.) around April 2013 seems to indicate that Ch.Supt. Ovens has introduced his own methods for recording crime figures in Harrow, which unsurprisingly has worked in his favour, giving the impression of a decrease in the absolute levels of most of the so called MOPAC 7 crimes. Clearly this situation is far from ideal, since Ch.Supt. Ovens appears to have taken his brief to focus effort on MOPAC 7 crimes as a challenge to concentrate solely on those crimes, and reduce them by whatever means possible, including statistical sleight of hand and the ignoring other “lesser” types of crime. Further, if each Borough Commander is using their own crime recoding methodology, then London wide figures for each borough will be incompatible with each other. Harrow’s crime figures for the MOPAC 7 crimes are tabulated below:


Since there is no official category for “Violence with Injury”, the total of the “Violence Against The Person” class was used here.

With the exception of Violence and Theft from the Person we can see that crime rates have mostly fallen, and this is most likely the cause of the recent article.

However, it’s important to compare the figures with sanction detections, which in the case of Robbery and Theft from Motor Vehicles have fallen 12% and 62% respectively. Also, with so much effort going into MOPAC 7 crimes, the largest sanction detection falls are as follows, which unsurprisingly are mostly non-MOPAC 7 crimes:

  • Theft/Taking of Pedal Cycles (-80.0%)
  • Fraud & Forgery (-75.5%)
  • Theft from Motor Vehicle (-62.4%)
  • Going Equipped (-40.0%)
  • Robbery – Business Property (-30.0%)

[Taking of Pedal Cycles and Going Equipped are historically low numbers]

Further, as mentioned above, it seems likely that some drug trafficking offences have been downgraded into drug possession offences, this presumably would mean that a sanction might be easier to obtain by the police, but at the same time serious criminals will be receiving reduced sanctions. This may be happening for other crimes as well, but from these figures it is difficult to say one way or another.


We need a thorough investigation into how exactly crime figures are being recorded in Harrow, and whether the new system is more lax than the previous method. Reporting of crime statistics needs to be standardized across London. Sanction detections for all classes of crime should be the measure by which a Borough is judged to have the highest/lowest levels of crime.

Much more attention needs to be focussed on rape and violence. But also fraud, which we are repeatedly told is one of the fastest growing crimes in the UK. Local crime should be targeted by absolute highest average numbers – so for Harrow, the seven pre-April 2013 targets should have been:

  • Burglary in A Dwelling (156.4 offences/month)
  • Theft from Motor Vehicles (134.8 offences/month)
  • Other Theft (130.3 offences/month)
  • Assault With Injury (80.2 offences/month)
  • Harassment (72.3 offences/month)
  • Common Assault (61.2 offences/month)
  • Personal Robbery (51.6 offences/month)

The police should be given more and better training in how to accurately record crime figures. Disciplinary action should be taken against Borough commanders found to be condoning the “fiddling” downwards of local crime statistics.

Copyright (c) Praxis Reform Research 2014. Used with kind permission. You can get in touch with @PraxisReform on Twitter. You can also download the original PDF of the research paper here, in it’s original, unedited, form.



(Visited 19 times, 1 visits today)

1 comment

  1. PraxisReform

    It must be slow news week at iHarrow, if Paul is publishing something so technical, such a long time after it was originally released… But, I’ll summarise here for those who want the executive briefing version.

    I’m a bit fed up with Ch.Supt. Ovens (and Councillors etc.) saying “[within a narrow and ill-defined range of crimes (or whatever else might be under scrutiny)] figures are down, didn’t we all do well?”

    The laws of chance tell you that even if Harrow’s police and Councillors did nothing else but sit about slurping coffee and eating doughnuts all day (I’m using this as a hypothetical case, not an insult) then some crimes would be up and some would be down — thusly, there’s always something to cherry pick after the fact and assert that it was their magic touch that made go up or down.

    What we should be hearing instead from Ch.Supt. Ovens is “Shootings are down, well done all concerned. However, beatings are up so we’ll be doing X, Y and Z to counter this trend.”

    In that way, we know:
    1. The Policeman / Councillor responsible for has noted the problems as well as good points
    2. There is some sort of plan of action going ahead to make improvements
    3. At the next review, we can see if doing X, Y and Z did what the person said it would

    So, put very simply, if someone (or an organization) can consistently identify problems, and can regularly make plans that usually (nobody is perfect) result in an improvement to the situation noted, then we have objective criteria by which we can say “this person (or organization) is doing a good job”.

    As far as I’m aware, none of this exists within Harrow’s Police or Council operations.

Comments have been disabled.