Overview of costs and benefits associated with regulation in Scottish agriculture

2. Overview of the assessment of cost and benefits of regulation and an assessment of the potential application of the Standard Cost Model in Scotland

2.1 Assessing the costs and benefits of regulation

In common with other policy areas, the assessment of the costs and benefits of agricultural regulation is an integrated part of the Impact Assessment ( IA) process. The IA process is outlined in the flowchart in figure 2.1 ( Appendix D).

The Better Regulation Executive provides guidance on various aspects of the IA process. Their guidelines on the assessment of costs and benefits are as follows (Better Regulation Executive 2008):

In essence an RIA is designed to examine a range of policy options in reference to a "do nothing" or "do minimum" option. Costs captured within an RIA will therefore represent the incremental cost of an option compared to this benchmark.

The guidance states that the summary must take account of the full range of costs and benefits: economic, social and environmental; and these should be monetised as far as possible. Other key non-monetised costs and benefits by main affected groups should be quantified or qualified on the summary page in cases where monetisation is not feasible. Major environmental impacts are specified as an example of where monetisation may not be feasible, and the valuation of non-market goods can indeed pose challenges.

Detailed guidance on how to assess costs and benefits is provided in the Treasury Green Book ( HM Treasury 2007), which gives advice on practical matters such as discounting and valuing costs and benefits where no market data exists. For example, figure 2.2 ( Appendix D) shows the Green Book approach to identifying the most appropriate valuation technique for non-market benefits.

In Scotland, the assessment of costs and benefits is undertaken by departmental policy experts as part of the Regulatory Impact Assessment ( RIA) process. The Scottish Government Improving Regulation Unit ( IRU) oversees the process, providing advice and auditing RIAs. The RIA process in Scotland follows the Better Regulation Executive guidelines, although the IRU provides supplementary guidance ("Scottish Extras") on IA (see Box 1 in Appendix D). The supplementary guidance covers the following points:

• Micro Business Test
• Legal Aid Impact Test
• "Test Run" of business forms
• Ministerial sign-off
• Review Regulatory Impact Assessments

2.2 Evaluation of the Standard Cost Model ( SCM) and its potential application to agriculture in Scotland

2.2.1 Overview of the Standard Cost Model

This section gives a brief overview of the SCM, for more detail see the UKSCM Manual (Better Regulation Executive 2005) and the International SCM Manual (available at: http://www.administrative-burdens.com/).

The SCM is a methodology for measuring the administrative cost of regulation imposed on business by government. It was developed in the Netherlands and is sometimes referred to as the "Dutch Model". In essence, it is a consistent way of breaking down the administration required to comply with a regulation into a range of manageable, measurable components (see figure 2.3, Appendix D). It should be noted that the SCM only "estimates the costs of a defined set of administrative activities" ( NAO 2007, p16); it does not measure the policy costs or any of the benefits of a regulation. Furthermore, it does not measure "'one-off costs' nor the 'financial costs' of complying with regulation, such as paying tax or licence fees" ( NAO 2007, p16).

Each regulation is broken down into information obligations ( IOs), data requirements ( DRs) and activities. IOs are obligations to provide information to the public sector and consist of one more DRs, each of which necessitates certain administrative activities. For example, a regulation may involve the IO of applying for a licence, which has data requirements such as: providing a description of the installation; providing details of emissions, and measures currently in place to monitor emissions. Each of these DRs will in turn require certain activities. The seven sets of standard activities are (Defra 2006b):

1. Familiarisation with requirements

2. Gathering and assessing relevant information/figures

3. Preparing figures (including calculating, presenting, checking and correcting)

4. Reporting (including written descriptions, copying, filing, distributing or submitting information/reports)

5. Making settlements or payments

6. Holding meetings

7. Inspections

The cost of these activities (from which the total cost of the regulation is derived) is calculated using the basic SCM formula:

Activity Cost = Price x Quantity

= (tariff x time) x (population x frequency)

For example, figure 2.4 shows the breakdown of the activities involved in applying to the regulator for a permit to operate an installation under the Pollution Prevention and Control Regulations 2000 (England and Wales). In this example, the cost of familiarisation (across all sectors, not just agriculture) is:

Activity Cost = (£14.06 x 22hrs22.37minutes) x (1610 x 1)

= £315.01 x 1610

= £507,166

A 30% overhead is added to these costs to give the internal admin cost for the activities. The external cost of the activities is the cost of the goods and services purchased to undertake the activity, e.g. in applying for a permit the business may employ consultants or accountants. The total admin cost is the sum of the internal and external costs.

The calculation in figure 2.4 ( Appendix D) distinguishes between admin cost and admin burden. The admin burden is the admin cost minus the cost of activities the business would have undertaken anyway (either for organizational reasons or in response to other regulations) in a business-as-usual ( BAU) scenario. The UK Administrative Burdens Measurement Exercise ( ABME), from which the example in figure 2.4 is taken, is unusual in making this important distinction and attempting to quantify BAU factors.

While conceptually simple, the actual SCM process is lengthy and requires significant resources. The steps involved are outlined in table 2.1.

Table 2.1 Steps in the measurement using the SCM (Better Regulation Executive 2005, p32; see http://www.berr.gov.uk/whatwedo/bre/index.html)

Phase 0: Start-up

Initial meetings of the department, Better Regulation Executive, consultants and other key stakeholders

Phase 1: Preparatory Analysis

Step 1: Identification of information obligations, data requirements and administrative activities and classification by origin

Step 2: Identification and demarcation of related regulations

Step 3 Identification of segments

Step 4 Identification of population, rate and frequency.

Step 5 Business interviews versus expert assessment

Step 6 Identification of relevant cost parameters

Step 7 Preparation of interview guide

Step 8 Expert review of steps 1-7

Phase 2: Time and cost data capture and standardisation

Step 9 Selection of typical businesses for interview

Step 10 Businesses interviews

Step 11 Completion and standardisation of time and resource estimates for each segment by activity

Step 12 Expert review of steps 9-11

Phase 3: Calculation, data submission and reports

Step 13 Extrapolation of validated data to national level

Step 14 Reporting and transfer to database

Detailed description of the SCM process, and assumptions underpinning it is beyond the scope of this brief overview. However, attention is drawn to the following key point:

"The SCM relies on deriving estimates of the standard cost of meeting each information obligation/data requirement ( IO/ DR) within a regulation for a 'normally efficient business'. Given the need to manage the overall costs of data collection while providing information about a very large number of information obligations/data requirements ( IO/ DRs), the SCM relies on the input of a very limited number of experts and/or businesses regardless of the number of businesses affected by each IO/ DR. As such, it does not produce a statistically representative measurement of costs: instead, it is a pragmatic approach to measurement that gives indicative estimates of the magnitude of costs. These estimates provide a starting point for setting reduction targets and highlighting the areas to focus upon. Moreover, the nature of the estimation process means that the greater the level of disaggregation of the cost estimates, the greater the potential margins of error surrounding the cost estimate." Defra (2006a, p11)

Variations of this caveat are repeated frequently in SCM literature, and it underlines a fundamental feature of the SCM: its strength lies in measuring (and demonstrating) change in admin costs over time, which is why it is often used in burden reduction initiatives. In the words of the Better Regulation Executive (2005, p7) the SCM "does not give a statistically representative measurement…instead it is a pragmatic approach to measurement that gives indicative data on the size of burdens to allow reduction targets to be set and areas to focus on identified"

In 2003 a group of European countries formed the SCM Network in order to foster a common approach to reducing regulatory burden. Table 2.2 shows the current members of the network:

Table 2.2 Current members of the SCM Network
Australia, Victoria	Ireland
Austria	Latvia
Belgium	Luxembourg
Czech Republic	The Netherlands
Cyprus	Norway
Denmark	Poland
Estonia	Portugal
Finland	Slovenia
France	Spain
Germany	Sweden
Hungary	United Kingdom
Italy

Of particular relevance to Scotland is the SCM work that has been undertaken as part of the UK central Government Administrative Burdens Measurement Exercise under the aegis of the Better Regulation Executive.

2.2.2 Application of the SCM in the Administrative Burdens Measurement Exercise

The ABME was launched in response to the Better Regulation Task Force's report "Regulation: Less is More". It was implemented between September 2005 and May 2006 in three phases:

1. Preparatory analysis - identifying regulations and breaking them down into constituent parts (information obligations, data requirements and activities)
2. Time and cost data capture and standardisation
3. Calculation, data submission and reporting

Sixteen Government departments took part in the exercise (including Defra), which covered all regulations in force in May 2005 (excluding tax and financial services, which were covered in separate studies). In total, the exercise studied 1400 regulations, 20,000 information obligations and 80,000 activities and required contacting 350,000 firms so that 8500 interviews could be undertaken. At its peak the ABME involved 700 external staff and 300 internal staff.

The findings of the Defra ABME are reported in detail in Defra (2006a). Selected findings are summarised in tables 2.3 - 2.5 ( Appendix D).

Key points are:

• Defra had (at May 2005) 362 regulations within the scope of the exercise and 3210 information obligations or data requirements ( IO/ DRs)
• the total admin cost to business for these regulations is estimated to be £735.7m (£2006)
• a small number of regulations and IO/ DRs are responsible for a large proportion of the cost, e.g. "the 20 most costly regulations in Defra account for about three-quarters of the total cost" see table 2.3 (Defra 2006, p13) and 61 IOs (out of 3210) accounting for about four-fifths of the total cost: "a comparatively small number of key activities are driving a significant proportion of the administrative costs falling to business" (Defra 2006, p13)
• the majority of the most costly regulations fall within the (a) animal health and welfare and (b) environment policy areas
• across all of Defra's regulations, 67% of the admin costs arise as a result of international (mainly EU) obligations, however this % varies considerably between different policy areas (see table 2.5, Appendix D); of the 67%, 46% provide no domestic discretion, and 21% provide for some domestic discretion in terms of which information businesses have to produce.
• admin cost v admin burden. An indication of how much of the costs are burden (i.e. costs incurred undertaking activities over and above those that would have been carried out anyway) is given by the % of businesses (for all industries covered by Defra regulations) that said they would have undertaken the data provision for some other organisational purpose. Two thirds (65%) said they would not have undertaken the data provision, while 24% said they would have "to some extent" and 11% would have "to large extent". . This implies that the admin burden is likely to be between 65% and 89% of the admin cost in the set of businesses surveyed.

While these findings are based on the UKABME, they provide useful insights because in many cases there is a great deal of similarity between the regulatory landscape in Scotland and the rest of the UK, despite the devolution of responsibility for agricultural and environmental policy. This is partly because about half of the admin burden of Defra's regulations are the result of EU obligations, which provide little scope for interpretation by the devolved administrations (see table 2.5). However, there are also likely to be differences in terms of the admin cost between Scotland and the rest of the UK in some policy areas, as a result of: (a) different regulations (e.g. in terms of bovine TB or land reform); (b) variations in the ways in which the same regulations are implemented (e.g. in terms of the license fees and approaches to enforcement); (c) differences in economic structure (e.g. predominance of grassland; nationalised water company); (d) differences in terms of demography and the physical environment (e.g. lower population density, high quality environment). So, while the overall findings are of relevance to Scotland, the admin costs of individual policies will be quite different in some cases. The issue of the transferability of costs from the ABME to Scotland is explored further in case study 5 ( PPC Regulations).

2.2.3 The SCM in practice: review of the performance of the SCM in the UK central Government ABME

Main purpose of the ABME/ SCM

A point made throughout SCM guidance (e.g. see Defra (2006a, p11) is that the main purpose of the SCM is to provide something to measure progress against because "it's important to show how things have changed (for the better)..." (Hammond 2008) if you want to demonstrate that you're serious about tackling regulatory burden - "...if businesses do not perceive the difference, then you really have failed". In order to demonstrate change in the admin burden it is essential to "get business involved in measurement and re-measurement" (Hammond 2008). This matters because simply achieving change may not be sufficient to allay industry concerns about regulation. Even though significant admin burden reductions have been achieved in Denmark and Holland (20%), they either (a) haven't been noticed by business as they have been spread thinly across many firms or, (b) when noticed, haven't been attributed to the admin burden reduction initiatives. Also new businesses tend not to notice reductions in burden as they are unfamiliar with the previous regulatory regime, so have no basis for comparison. It seemed to Hammond (2008) that changing perceptions of regulatory burden is as difficult, if not more so, than actually reducing the burden, and that despite the Better Regulation initiatives and the relatively light touch regulatory regime in the UK "perceptions are still negative". In this light, current stakeholder engagement with, and support for, SEARS is both significant and welcome. The Scottish Government intends to commission research to establish customer's experiences of interacting with SEARS.

Practical Issues

There are a huge number of activities to measure, which makes it difficult to break down all tasks by firm size and industry sector - this level of segmentation would make the exercise unmanageably large. Also it can be difficult to get a standard cost for some activities, especially from interviews. There can be a great deal of variance in the costs, depending on how people interpret the questions. It proved particularly difficult to get hold of people to interview in certain industries, notably fishing and agriculture, where respondents were often unavailable and/or uninterested. This problem can be tackled by tailoring the data collection, e.g. using out of hours calls, and avoiding busy times such as harvest or lambing. Existing data collection mechanisms such as the Farm Accounts Survey ( FAS) could also be used, as could information gathered through SEARS - although it is acknowledged that the latter has tended to gather qualitative rather than quantitative data.

Hammond (2008) emphasized the "need to put it (the questions about time) in ways people understand" and suggested that "expert panels are one of the most effective ways of getting information". These work by presenting a "straw man" (a scenario outlining what you think would be involved in admin for a regulation) to a panel of industry representatives for discussion. This proved to be less time consuming than phone interviews. Another way round the problem of segmentation is to identify the size of most of the firms; if most are small then focus resources on this segment (most farms will be micro (0-9 employees) or small (10-49) by the SCM manual definition). Another reason for focusing on small firms is that they are often time constrained and tend to find admin problematic.

Advantages

According to Hammond (2008) there are two big advantages to the ABME/ SCM approach. Firstly, "it gives you a good understanding of the regulatory landscape", e.g. it helps you to identify where the same information is being collected multiple times, or where admin is particularly problematic. Secondly, "it helps to put momentum behind the simplification measures". The NAO (2007, p7) concurred with this assessment, observing that (a) the ABME may deliver "a better understanding among policy officials of the regulatory regime and how it affects business" and (b) that "setting targets has provided a new focus and impetus to Government efforts to reduce the administrative burdens of regulation". The impetus for change is important as the impact of the ABME/ SCM depends on the scope that exists for pursuing change in light of its findings.

2.2.4 Cost and Cost-Effectiveness of the ABME/ SCM

An SCM based ABME involves considerable resources. Full costings for the ABME were not available, in part due to the difficulty of estimating the costs of the internal resources used in the exercise. A preliminary appraisal has been carried out, using the best available data (see table 2.6, Appendix D). The results indicate that the UK central Government ABME/ SCM (across all industries and Government departments apart from HM Revenues and Customs) will have a cost of approximately £21.2m ( NPV, £2006) over its five year lifespan. It is estimated that a full ABME would cost approximately £2.7m to apply to all RERAD regulations (assumed to be 181 regulations, across all industries). Preliminary evidence suggests that the cost reductions achieved by the ABME in UK central Government are significantly greater than the costs (B:C ratio of 31). However, it should be noted that these conclusions rest on a series of assumptions and are therefore somewhat uncertain. For example, this result is based on the assumption that all the savings in reaching the 25% admin cost reduction target by 2010 can be attributed to the ABME. The ABME has clearly been a vital part of the drive to reduce admin costs. Nevertheless, a full feasibility study should clarify the validity of these assumptions and undertake rigorous, transparent costing.

Undertaking an SCM-based ABME is not the only way to improve regulation. The valuation of benefits is frequently missing from Impact Assessments, and better valuation is perhaps another way in which some regulations could be improved. The cost of the ABME per IO is £1061, which compares favourably with the resources required for valuation. In some ways, this is not a valid comparison as an ABME and valuation answer different questions; the former identifies less costly ways of implementing a regulation, while the latter addresses the question of whether or not a regulation should be implemented at all. Nevertheless, given that adherence to the official guidance for IAs appears to be somewhat variable, it would perhaps be helpful if it were followed more rigorously or - where not - some explanation were offered as to why not. For example, in some cases, the costs of a valuation exercise may be disproportionate to the additional insights gained or of little operational relevance if compliance with an external standard is required under an international obligation.

2.2.5 Reducing the costs of implementing an ABME/ SCM approach in Scotland

The estimated cost of undertaking an ABME in Scotland (see table 2.6, Appendix D) should be considered a high estimate as there is considerable scope for these costs to be reduced. As Defra (2007b, p3) note:

"while administrative burden measurement is useful, it is not necessary for all countries to undertake such a resource-intensive exercise to adequately inform simplification programmes, providing there is a representative sample of data for specific policy areas."

There is a sizeable body of secondary data that could be used to cut down the amount of primary gathering required in Scotland (see table 2.7 for potential sources). Data already exists on the structure of farming in Scotland, and now that an SCM-based ABME has been undertaken by UK central Government, there is a large amount of data on the information gathering activities required for Defra regulations, which could be adapted for use in a Scottish ABME. The data is available in a web-based calculator format (see figure 2.4, Appendix D) or as an Access database. An updated version of the Access database will be available from the Better Regulation Executive in September 2008. An examination of the limitations of using UKABME data to calculate the admin costs of regulations in Scotland in given in case study 5.

Another way to reduce the costs is to have a targeted approach, i.e. apply the method in key policy areas, such as areas where admin costs are likely to be a particular problem. According to Hammond (2008), this approach has been adopted in France, and the Irish Government is also considering doing a cut down ABME. This is understandable when one considers that a relatively small number of regulations are responsible for the bulk of the admin burden (see table 2.3, Appendix D). It makes sense to focus on these regulations and identify the causes of the costs. If the complexity of the regulatory data requirements is driving the costs (as opposed to the sheer number of businesses affected), there may well be a case for using the ABME/ SCM approach.

Table 2.7 Potential Secondary Data Sources for a Scottish ABME

Data required	Potential sources
Identification of IOs/ DRs	Defra ABME, BRE Admin Burdens Calculator, 2008 SEPA Survey of Regulated Industries?
Quantification of time taken per IOs/ DRs	Defra ABME, BRE Admin Burdens Calculator, 2008 SEPA Survey of Regulated Industries?
Internal costs (e.g. farm wage rates)	Defra ABME, BRE Admin Burdens Calculator, Agricultural Census, Farm Accounts Survey
External costs (e.g. consultants)	Defra ABME, BRE Admin Burdens Calculator, Agricultural Census, Farm Accounts Survey
Population data	Agricultural Census, Farm Accounts Survey, Scottish Pollutant Release Inventory

2.3 Conclusions

The ABME/ SCM approach to better regulation has significant strengths, not least of which are that it: demonstrates commitment to reducing regulatory burden; improves the regulator's understanding of the regulatory landscape; provides momentum for simplification plans; enables comparison of changes in administrative burden, through time and between government departments and countries.

The preliminary appraisal of the ABME/ SCM approach indicated that it may well be a cost-effective exercise. However, in practice, the effectiveness of the approach will depend very much on how it is implemented and the policy aims that one is trying to achieve. Humpherson (2006) warns of the dangers of misunderstanding the purpose of the SCM:

"The great allure of the SCM is its appearance of scientific objectivity and of (largely illusory) accuracy. The key thing is not so much that regulation y cost £x, but that the whole regulatory system is targeting reductions and seeking to achieve simplifications."

The SCM is not a way of measuring administrative costs more accurately, it is primarily a way of being able reliably to repeat measurements over time and demonstrate change, i.e. it is a technique that provides cost estimates of low validity but high reliability. It is therefore most effective when employed as an integrated part of an ongoing programme of admin cost reduction, rather than as a one-off assessment of costs. In this context, the overlap between existing SEARS activities and any potential SCM analysis would merit consideration. That is, significant effort has already been expended under SEARS to identify and reduce burdens and this information could inform - or possibly make redundant - a generic or agricultural SCM exercise.

Recommendations

1. It is unnecessary to fully replicate the ABME/ SCM exercise undertaken in UK central Government as many lessons have already been learned and do not to be re-learned, e.g. data gathering techniques should to be tailored to the respondents in terms of timing, format, question wording etc. There is considerable scope for reducing the cost of an ABME/ SCM exercise through targeting and the use of existing data sources.
2. Any Scottish exercise should focus on key regulations (i.e. high admin cost and irritation, high complexity, Scotland specific) as a disproportionate proportion of the administrative burden arises from a small number of regulations. Sharper focus would enable a Scottish exercise to complement, rather than replicate the UK exercise by, e.g., improving the validity (larger samples) and granularity (estimating standard costs for smaller segments of industry) of the standard costs. There is considerable scope for reducing the cost of an ABME/ SCM exercise through targeting and the use of existing data sources. Information and experience gained through SEARS will be highly relevant to such an exercise. Indeed, it may be that SEARS already offers more agriculturally relevant insights and remedies than a separate SCM exercise would, even though the latter might be more quantitative in nature.
3. The UKABME/ SCM will be revised/repeated in 2010; this represents a significant window of opportunity for partnership.
4. Achieving reductions in admin burden on their own are unlikely to improve industry confidence in regulation. Reductions need to be clearly demonstrated to the industry, not just achieved, and this requires the close involvement of the industry in the measurement and re-measurement process.