On the Relationship between RBKC Civil Enforcement Officer Penalty Charge Notice Issue Rate, Overtime Allocation and Prospects of Continued Employment

Introduction

Anecdotal evidence suggests that Civil Enforcement Officers (CEOs) working for the Royal Borough of Kensington and Chelsea (RBKC) may be given the opportunity to do more overtime work if the rate at which they issue Penalty Charge Notices (PCNs) is high. Conversely, CEOs issuing PCNs at a low rate may be given less overtime work, and may in fact be at substantial risk of having their employment terminated. Interested readers may wish to read more about these issues which were highlighted at the employment tribunal of former RBKC CEO Hakim Berkani.

Despite the enduring topicality of parking enforcement as a subject of UK-wide interest and debate, quantitative studies which explore the reliability and consistency of the anecdotal evidence described above have, to the best of this author's knowledge, not been carried out. In this context we carry out statistical hypothesis tests to establish whether there is in fact (a) a significant positive association between high rate of PCN issue and allocation of overtime and (b) a significant association between low rate of PCN issue and termination of employment.

Data Set

Under the Freedom of Information Act, on 2 January 2013 information was requested from the Royal Borough of Kensington and Chelsea regarding Parking Enforcement as carried out by NSL on behalf of RBKC in the 2012 calendar year, broken down by Civil Enforcement Officer (CEO). Specifically, for each calendar month of the year 2012 and each CEO, RBKC was asked to provide the following statistics:

  1. Anonymous Identifier of CEO (can be e.g. a badge number or any other numbering, as long as the same CEO gets the same identifier from month to month)
  2. Number of PCNs issued during the month by CEO
  3. Number of hours (including overtime) worked by CEO
  4. Number of overtime hours worked by CEO

This FOI request was responded to by RBKC on 7 February 2013 giving data for 111 CEOs. The full original data set is available (as supplied in PDF format), as well as an Excel spreadsheet of the data and its subsequent analysis.

Method and Results

The number of PCNs issued per hour for the 2012 calendar year was computed for each RBKC CEO, and the Excel function RANK.AVG was used to assign a rank to each CEO according to this metric. The CEO issuing the least number of PCNs per hour (rank 111) was CEO KC8 who issued 0.276 PCNs/hr. The CEO issuing the most PCNs per hour (rank 1) was CEO KC85, who issued 1.355 PCNs/hr. The rank assigned in terms of this metric will be called the "Performance Rank".

The RANK.AVG function was similarly used to rank the CEOs according to the number of overtime hours they had been allocated. The rank assigned in terms of this metric will be called the "Overtime Rank".

A Spearman's rank correlation coefficient between the Performance Rank and the Overtime Rank was computed in two different ways, yielding in both cases a strong positive correlation of 0.527. The p-value associated with this result (i.e. the probability of such a strong correlation occurring by chance if there was no association between the Performance Rank and the Overtime Rank) is less than 0.001. Thus there is a strong, statistically significant, association between Performance Rank and Overtime Rank. From the CEO perspective, it is hard to escape the impression that issuing more PCNs/hr is a good way to have more overtime allocated.

Likewise for each CEO it was examined whether or not they were employed in December 2012 (i.e. had a number of working hours for December 2012 greater than zero). This metric was called "Employed at Year End?" and was set to 0 (not employed in December 2012) or 1 (employed in December 2012) for each CEO. Interestingly, of the 25 poorest performing (in terms of PCNs issued per hour) CEOs, only 9 were employed at year end, while of the 25 highest performing, all 25 were still employed at year end. The Pearson correlation coefficient between the "Employed at Year End" and "Performance Rank" is positive and highly significant (r=0.499, p < 0.005). That is, the probability of such a strong correlation occurring by chance if there is no association between Performance Rank and continued employment, is less than 0.005. We can therefore conclude that there is in fact a strong, statistically significant, association between Performance Rank and continued employment. From a CEO perspective, it is hard to escape the impression that a poor rate of PCNs/hr puts ones employment at risk, while a good rate of PCNs/hr is likely to lead to continued employment. We note that in the absence of reasons for termination of employment more convoluted explanations are also theoretically possible e.g. CEOs who do not like their jobs are probably more likely to resign after a short period of employment that featured a low rate of PCN issue.

In order to gain more information about the procedures used to allocate CEO overtime, a follow up FOI request FOI 2013-152 was made on 8 February 2013. This was responded to by RBKC on Thursday 14 March 2013. The information sought and the corresponding replies were as follows:

  1. Q: Who makes the decision about the allocation of overtime to a CEO?

    A: The CEO's NSL supervisor

  2. Q: What factors are typically taken into account in determining the quantity of overtime allocated to a CEO? (e.g. financial need, high PCN issue rate, punctuality, neatness, low PCN cancellation rate etc.)

    A: Supervisors will seek volunteers who produce good quality work i.e. no errors in tickets, have low sickness levels, are experienced to cover the area needed and have good overall conduct.

  3. Q: At what rate are CEO overtime hours paid relative to CEO regular working hours?

    A: CEOs are paid overtime at the normal hourly rate except for Sundays (paid time and a half) and Bank holidays (paid double time)

It is left as exercise to the reader as to whether the reality disclosed in the 2012 data analysed above is consistent with the Council's reply to Question 2 above.

Discussion and Conclusion

Parking is big business for many Councils in the UK. RBKC made a surplus from its on-street parking operations of GBP 27 million in 2011/2012 (up from GBP 20 million in 2010/2011). While the law nominally forbids parking to be used for revenue raising purposes and parking contracts therefore no longer contain explicit annual PCN, clamping and removals targets as Key Performance Indicators, it seems on-the-ground operating practices have been slower to adjust. As shown above, the reality for CEOs on the ground is that there is a strong association between the number of PCNs they issue per hour and their corresponding overtime allocation. There is also a strong association between the number of PCNs they issue per hour and their prospects of continued employment. CEOs live in a world consistent with the hypothesis that, if they want more overtime and if they want to continue to be employed, then they should issue more PCNs per hour. That Councils may turn a blind eye to the sharp practices their CEOs are subjected to in this regard is perhaps not surprising given their apparent primary concern is the protection of their parking revenue rather than the health and safety of their CEOs. But perhaps it is time to introduce legislation making it illegal for Councils to operate in any manner that leads to a statistically significant association between CEO benefits/remuneration and CEO revenue generation. In this context it may be worth examining the latest contracts between Councils and CEO operators (such as NSL in the case of RBKC) which feature metrics such as "Street Visit Percentage", which would seem to naturally correlate with revenue generated by PCN issue, and which are tackled with such vigour that they are frequently way-over-achieved (285% in 2011/2012).

IMPORTANT NOTE: The views expressed herein are those of the author and not of Imperial College London. In interpreting these findings is important to understand the difference between correlation and causation. The primary purpose of this article is to present evidence of the existence of two statistical correlations which are extremely unlikely to have arisen by chance. In this face of the evidence above, it would be a brave person who would make a claim that the correlations found do not actually exist. Naturally, however, the precise reasons for why the correlations have arisen, whether or not there is a direct causal link, and what confounding factors may be at work is open to debate and Councils and motorists are unlikely to present the same arguments in this regard. Follow-up FOI requests for details regarding e.g. reason for termination of employment, whether termination of employment was voluntarily or involuntary, and the number of disciplinary hearings to which each CEO was subjected during the year would help to clarify the issues further.

William Knottenbelt
April 2013, updated August 2013

Back to my parking page