Regional variability in the application of JSA sanctions

Variability of JSA Sanctions between JCP Districts

An investigation into the level and appropriateness of JCP’s use of benefit sanctions, including differences of approach between JCP Districts

Submitted to the House of Commons Select Committee to examine the role of Jobcentre Plus in the reformed welfare system.

By TheUxbridgeGraduate

An independent analyst

Prepared in April 2013

Specific Terms of Reference

To comment on JCP’s role in relation to the rights and responsibilities of benefit claimants, including: the effectiveness of benefit conditionality, particularly job-seeking conditionality and the mandatory “work-focused interview”; and the level and appropriateness of JCP’s use of benefit sanctions, including differences of approach between JCP Districts.

Section 1   Executive Summary

 My submission focuses on the final sentence in the Specific Terms of Reference, namely,“the level and appropriateness of JCP’s use of benefit sanctions, including differences of approach between JCP Districts”.

  1. In March 2013, I conducted a brief statistical study to ascertain whether JSA sanction rates varied significantly across England. This was in response to concerns that targets were driving JSA sanctions. The study was undertaken on the belief that if a single, over-arching target was in operation across the UK then variation between regions would be found to be random and statistically insignificant. That is, sanctions rates would be essentially uniform across the UK.
  2. My study found statistically significant variation in sanctions rates between the eight English areas selected for the study.  In particular, sanctions regimes appeared statistically significantly harsher in London and the West Midlands than in Tyne and Wear, Merseyside, and the other northern metropolitan counties.
  3. A conclusion from that study was that a single national target did not seem to be in operation. However, the conclusion was tentative because of the imperfect data used in the study.
  4. I have since undertaken a fuller and better designed statistical study on the same eight areas. This second study forms the platform upon which this submission is based. I believe the results from the second study are reliable.
  5. The latter study confirms the first study’s conclusion, i.e., that a single national target is not in operation, that Tyne & Wear and Merseyside have the lowest sanctions rates, that London and the West Midlands have higher than expected sanctions rates.
  6. Both studies suggest that sanctions rates are a policy variable under the control of managers at regional, district or job centre level. Hence arbitrary and aggressive sanctions decisions may be the product of delegated sanctions policies, resulting in a post code lottery for claimants.
  7. The Select Committee is urged to seek an explanation from DWP for why sanctions policies are more aggressive in London and West Midlands. The Select Committee is also urged to consider whether aggressive sanctions policies are compatible with the social security system, a large part of which is administered by DWP.

Section 2   Background

  1. There have been several disturbing recent accounts of JSA claimants being sanctioned for seemingly trivial reasons. A sanction is usually a very serious penalty to impose on an individual for whom JSA may be the sole source of income. Liberal application of sanctions runs counter to the notion of social security which ostensibly is to provide a safety net to people in need. Removing an individual’s benefit can realistically propel that individual into destitution. One would therefore hope that a sanction would be applied only as a last resort when all other measures have failed.
  2. Some commentators have suggested that aggressive sanctioning of claimants has arisen because targets have been set, either at the centre, or by autonomous regional or district managers, or even by individual job centre managers themselves. Some compelling anecdotal and written evidence has recently emerged that may confirm the existence of regional or district targets.

 Section 3    Scope of my study

  1. Initially, the purpose of my investigation was to ascertain whether sanctions were being applied to meet targets. By establishing whether significant variation existed between regions, I hoped to conclude one way or another whether targets were guiding sanctions use.
  2. Because of data and possible methodological limitations, I have now confined my analysis to simply ascertaining whether significant regional variation exists in sanctions rates or whether sanctions rates are uniform across the regions.
  3. Confirmation of uniform sanctions rates across regions admits of one of the following conclusions set out in points 4 and 5:
  4. A nationally set target exists and hence uniform sanctions rates arise through individual job centres complying with instructions received from the centre in pursuit of the target.
  5. Uniform sanctions rates arise solely because claimant job-seeking behaviour is uniform across regions. In this case sanctions solely reflect appropriate responses by job centre staff/decision makers.
  6. Confirmation of differential (non-uniform) sanctions rates between regions admits of one of the following conclusions, set out in points 7 and 8
  7. Claimant job-seeking behaviour varies from region to region and the application of differential sanctions rates    reflects appropriate responses by job centre staff/decision makers.
  8. Sanctions policy is being determined at regional or district level, or even by individual job centre managers. If true, then a    post-code lottery is faced by claimants who may be subject to an arbitrary and aggressive sanctions policy in one region or to a more nuanced policy in another.
  9. My main purpose is to provide numerical and statistical evidence to the Select Committee on this topic. I also set out the values which I hope will guide the Select Committee in its deliberations.
  10. All data used in the study have been set out in tables in later sections with the sources (ONS and DWP) acknowledged.

Section 4   Methodology

  1. The study rests on the assumption that the number of sanctions issued will be proportional to the number of claimants on a region’s books, i.e., higher claimant counts will result in higher sanction counts.
  2. The six English metropolitan counties, plus Inner and Outer London, were selected for the study on the ground of homogeneity. The number of regions was restricted to eight to make the data collection and processing more manageable. Including more areas would not add much additional information to confirming whether sanctions rates are uniform or not.
  3. The claimant counts for each of the eight regions for the period 10 May 2011 to 10 April 2012 were collected from the ONS website. The average monthly claimant count for each of the eight regions was computed. These average monthly claimant counts were used to weight and distribute the actual sanction count over the regions to obtain an expected sanctions count under the hypothesis that sanction rates are uniform.
  4. The actual sanctions counts compiled by DWP for the period 1 April 2011 to 21 October 2012 for each of the eight regions were compared to the expected sanctions count.  The sanctions data supplied by DWP were not broken down into months and so a time-series analysis was not possible.
  5. The actual and expected sanctions counts for each region are shown in the following table, together with the average monthly claimant counts used to calculate the expected counts. A visual plot of actual versus expected sanctions counts also follows.

Region

Period 11 May 2011 to 10 April 2012

Period 1 April 2011 to 31 March 2012

Monthly claimant count on average

Sanctions count

Count (ONS)

%

(Calculated)

Actual

(DWP)

Expected

(Calculated)

Tyne and Wear

39,186

6.20%

9,770

12,202

Merseyside

50,983

8.07%

14,030

15,876

South Yorkshire

43,363

6.86%

13,330

13,503

Greater Manchester

82,450

13.05%

25,600

25,675

West Yorkshire

70,267

11.12%

22,010

21,881

Inner London

111,548

17.66%

35,870

34,736

West Midlands

111,925

17.72%

36,150

34,853

Outer London

121,944

19.31%

39,940

37,973

Total

631,666

100.00%

196,700

196,700

Sanctions for report good version

6.   Under the hypothesis of uniform sanction rates across the regions a test statistic was computed and compared to the critical values of the Chi-squared distribution with seven degrees of freedom.

7.  The value of the test statistic is 889.89, which is a very high and emphatically confirms that sanction rates are not uniform across regions. For interested readers, here are the workings for the computation of the test statistic.

Sanctions

Diff

Diff squared

Diff squared / E

Actual

Expected

Tyne and Wear

9,770

12,202

-2,432

5,916,443

484.86

Merseyside

14,030

15,876

-1,846

3,408,286

214.68

South Yorkshire

13,330

13,503

-173

30,022

2.22

Greater Manchester

25,600

25,675

-75

5,613

0.22

West Yorkshire

22,010

21,881

129

16,637

0.76

Inner London

35,870

34,736

1,134

1,286,301

37.03

West Midlands

36,150

34,853

1,297

1,681,708

48.25

Outer London

39,940

37,973

1,967

3,868,189

101.87

Total

196,700

196,700

0

Test statistic

889.89


Section 5    Discussion of results

  1. Inspecting the above table enables us to see which regions are contributing to the very high test statistic of 889.89.  It is clear that Tyne & Wear, Merseyside, Inner London, West Midlands, and Outer London account for 99% of the test statistic’s value. The remaining three regions account for just 1% between them.  It is the five “high value” regions which should, because they are exceptional, attract attention.
  2. The study indicates that claimants in the two London areas and in the West Midlands are significantly more likely to be sanctioned than their peers in the other areas. Why this is the case may be worthy of further investigation. Either the extra sanctions arise because claimants in these areas are less diligent in their job searches, or managers of these areas are pursuing aggressive sanctions policies. Are area managers authorised to determine their own sanctions policy?  An aggressive sanctions policy might well cause aggrieved claimants to believe they have been sanctioned because targets are the driving force.
  3. In contrast, claimants in Tyne & Wear and in Merseyside are less likely to be sanctioned than their peers in the other areas. Again, further investigation may be needed to explain this variability. Have the managers in these areas been told to “go easy” on claimants by the centre? Or has the more relaxed sanctions policy been set at regional level?

Section 6     Conclusions, summary and recommendations

  1. The statistics show overwhelmingly that the likelihood of a sanction being applied depends on region. The most likely explanation for this is that sanctions policies are being determined locally.
  2. This may raise issues of equity and fairness. For example, is it fair that a claimant should face a greater risk of a sanction simply by virtue of their area? Is it fair that a claimant’s technical or minor breach of a job seeker agreement is punished with a sanction in one area but is forgiven in another? Should not all claimants be subject to the same disciplinary regime irrespective of their region? I urge the Select Committee to consider this matter in its deliberations.
  3. Delegated sanctions policy may also bring forth an issue of accountability. The legislation empowers sanctions to be applied in pursuit of a specific purpose. Aggressive sanctions policies give the impression that some regions may be using sanctions for purposes other than authorised by law. For example, if area managers have taken it on themselves to use sanctions as a deficit reduction tool then arguably this is unlawful. Similarly, if these same managers have set targets for their regions then this is likely to be arbitrary and again unlawful. I suggest the issue of accountability and lawfulness may be relevant considerations for the Select Committee.
  4. Further to the issue of accountability, it seems unlikely that DWP can be unaware of the differential (non-uniform) sanctions rates across the regions. My guess is that DWP must be monitoring how sanctions are applied across areas, perhaps using methodologies similar to the one outlined in this submission. If so, then DWP should be in a position to explain and justify the differential sanction rates across regions I urge the Select Committee to seek such an explanation and justification from DWP.
  5. If DWP does not have systems to monitor how sanctions are being applied across regions then I would suggest its management control systems are remiss. I urge the Select Committee to explore this matter with DWP when its representatives appear.
  6. I also ask the Select Committee to bear in mind the purpose of social security when it considers sanctions in relation to job seekers.  An aggressive sanctions policy, where it exists, runs counter to the idea underpinning social security. I suggest it is incongruous to punish job seekers with further impoverishment, if not destitution, with sanctions imposed by a major component of the social security system. 

Section 7   Regional claimant counts (sourced from ONS)  (Redacted in on-line version) 

Section 8    About Me

I am a freelance, independent analyst. I trained as a Management Accountant and have received post-graduate training in industrial statistical methodology from a reputable British university. I am not in receipt of out-of-work benefits or under sanction.

My first study can be found via this link:

https://theuxbridgegraduate.wordpress.com/2013/03/23/are-uk-unemployment-benefits-being-unlawfully-withheld-from-claimants/

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s