Legal Aid, Sentencing and Punishment of Offenders Bill

Memorandum submitted by Citizens Advice Bureau (LA 54)

Response to the Public Bill Committee

At the evidence session on 14 July Damian Hinds MP requested further information from me in respect of our "business case for legal aid" figures, having noted the findings we published on this last year. I am more than happy to comply.

Let me start with the background to this research. It originated in discussions on Management Information issues which took place two years ago at senior level between Citizens Advice and the Legal Services Commission. These led to common agreement that we should find ways of analysing outcomes data collated in our respective case recording and research systems to inform policy development on future spending decisions and priorities for commissioning. Subsequently, our policy team was tasked to work with the LSC to look at case outcomes data with a view to developing some research, as a joint project between ourselves and the Commission, to explore the possibilities of designing empirically valid measures of cost-benefit for different types of legal aid

After looking at various possible approaches with colleagues in the LSC, the Legal Services Research Centre and Professor Roger Bowles, we decided that the way forward would be to put together an exploratory research paper - hence the title "towards a business case" - that we could present to the LSRC's international research conference as a discussion paper to spur further research into an area of analysis that had been hitherto unexplored.

The datasets which we drew from are statistically significant; the study looked at

· all the 2007/8 data for the civil and social justice surveys on problem incidence and 'adverse consequences' (a sample of around 6,000) undertaken by LSRC, and

· all the 2008/9 data on controlled work case outcomes and billed case data from all LSC contracted providers at the 'Legal Help' level in debt, welfare benefits, housing and employment (around 400,000 cases).

However, we recognised that there was a big problem with methodology and assumptions - especially in numerically identifying the proportions where legal advice does not help and cannot therefore make any saving - and the measurements used to quantify those savings.

We therefore built in a control category between those cases which experience adverse consequences with known costs and those that do not (LSRC survey data) and between those legal aid cases which record a beneficial outcomes and those which did not receive a beneficial outcome (LSC outcomes data). The methodology was specifically designed to strip out this "deadweight" from the analysis so that the results could not be seen as skewed.

The cost-benefit analysis algorithm arrived at was developed within the LSC, based initially on a worked example in the housing category. However, after being stripped of its policy function in early 2010 the LSC withdrew from the project. We decided to complete the paper on our own and the LSC continued to give us access to their data for this purpose.

The resulting business case paper was therefore very much a first stab at doing a cost benefit analysis of legal aid work, and we welcome you probing the figures. To answer to your two specific questions, I would say the following: -

1. Whether we have done any comparable research for non-legal aid channels

Yes. Quite separately from the analysis of LSC outcomes, Citizens Advice have also provided some analysis to BIS of savings to Government from CAB casework; this is also dealt with in the business case paper at paragraphs 93-96 and Table 10. These estimates were included in the paper as they are relevant and comparable.

2. Whether it is legitimate to represent loss of employment costs as GDP loss

This is a good question. We based adverse consequence costs on the analysis of the civil justice survey data in the LSRC’s paper "Mounting problems: Further Evidence of the Social, Economic and Health Consequences of Civil Justice Problems. [1] This found that loss of employment amongst survey respondents as result of legal problems resulted in direct costs to the public purse through benefit claims. For respondents claiming unemployment related benefits, the average period was 19 weeks, making the average costs of job loss £1,075. The cost at today’s values is likely to be higher, however, as the Mounting Problems research used the Jobseeker’s allowance rate of £55.65 applicable to April 2004 for a lone parent above the age of 18 or a single person over the age of 25.

In addition to these costs, loss of employment caused a net social cost measured in terms of lost output. The value of output foregone was measured using GDP per head, given the average spell of unemployment reported among respondents who lost their job as a consequence of a problem. The average value of lost output amounted to £8,140 – again this is an under-estimate as the figure was derived from average GDP per head on the basis of market prices pertaining in 2003, with a weekly value arrived at by dividing the annual value by 52.

On balance, we thought it was legitimate to factor in both sets of indicators from Mounting Probems for lost employment costs into the overall cost benefit analysis. The data on positive outcomes was then mapped over these adverse consequences. So for each category of law the model analysis does the following:

· Shows the total number of clients seen by legal help providers;

· Multiplies clients seen by percentage of adverse consequences;

· Analyses those numbers by the percentage that benefit from legal aid;

· Multiplies the numbers for each adverse consequence by the stand costs for those adverse consequences;

· Shows the total cost of provision for that category of law;

· Undertakes a sum of the adverse consequence avoided;

· Takes away cost of provision from cost of consequences avoided;

· Expresses total expenditure as a percentage of costs avoided.

The high level source data is contained in the appendicies to the paper; if you require the original spreadsheets we can supply these also. We do not claim that the methodology is perfect, or that is not open to challenge, especially about what should or should not be included in calculating the adverse consequence costs. However, we do think that the methodology is a good starting point for analysis and debate.

For further research on cost-benefit analysis in respect of legal aid spending, you might also be interested in looking at the following the following:-

On debt advice see

From Law Centres

From other jurisdictions see

Price Waterhouse’s research for the Australian system

Perryman research in Texas

Report from Canada BC

I hope the Committee finds this useful and answers the questions Mr Hinds raised.

July 2011

[1] Published in Transforming Lives: Law and Social process LSRC 2007

Prepared 7th September 2011