Landman Economics – Written evidence (GAM0039)

Written by Howard Reed (Director, Landman Economics)

Introduction

Landman Economics is pleased to submit this response to the call of evidence issued by the House of Lords Select Committee on the Social and Economic Impact of the Gambling Industry. Since 2013 Landman Economics has conducted research for the Campaign for Fairer Gambling on the social and economic costs of gambling, focusing initially on the impact of Fixed Odds Betting Terminals (FOBTs – also known as B2 gaming machines) and, more recently, online gambling. This submission summarises our main findings, and is particularly relevant to Questions 5 (“what are the social and economic costs of gambling?”) and 6 (“what are the social and economic benefits of gambling?” in the call for evidence. With reference to these questions, this submission focuses on specific aspects of the economic impact of gambling, as follows:

1. the links between gambling activity and measures of economic prosperity (household income and deprivation);
2. The relationship between the location of LBOs (licensed betting outlets – i.e. betting shops) and deprivation by local area;
3. The wider economic impact of FOBTs and online gambling – in particular the impact on employment and tax receipts.
4. The proportion of profits from the online gambling sector which arise from gambling activity by players who are defined as ‘problem gamblers’ or ‘at-risk gamblers’.

At the end of the submission I also address Question 8 (“how might we improve the quality and timeliness of research in the UK?”)

Page 10 of this submission is a one-page summary of the main findings.

It should be noted that all of the research findings in this submission are based on data that was collected before the recent reduction in maximum stake for FOBTs from £100 to £2 per play in April 2019.

Response to Questions 5 and 6

1               Gambling, income and deprivation

1. The 2016 Health Survey for England contains data on household incomes (adjusted for family size) and on the extent of deprivation in the area which each respondent lives in (using the Index of Multiple Deprivation, which is a measure of the extent of deprivation in each local area in England)[1]. Table 1 below uses data from the Health Survey for England 2016 to show the proportions of problem gamblers (both for any kind of gambling and separately for those who participated in online and FOBT gambers) according to the quintile of the household disposable income distribution (from the poorest 20 per cent to the richest 20 per cent of households, adjusting incomes to take account of family size). Table 2 shows the same information for respondents broken down by quintile of IMD (Index of Multiple Deprivation) from the most deprived areas up to the least deprived areas.

2. Tables 1 and 2 show that there is a ‘household income gradient’ as well as a ‘deprivation gradient’, in that poorer households are more likely to contain problem gamblers than richer households, and households in more deprived areas are more likely to contain problem gamblers than households in less deprived areas. The gradient is steeper for deprivation than for income: Table 2 shows that 1.5 per cent of survey respondents in the most deprived quintile are problem gamblers compared to 0.2% in the least deprived quintile. This means that adults in households in the most deprived quintile are more than seven times more likely to be problem gamblers than adults in households in the least deprived quintile. For problem gamblers who play FOBTs, the difference between the most and least deprived households is even more stark; adults in the most deprived quintile are ten times more likely to be problem gamblers than adults in the least deprived quintile. For online gambling the deprivation gradient is not as severe but people in the most deprived quintile are still four times more likely to be problem online gamblers than people in the least deprived quintile.

3. The analysis by disposable income quintiles in Table 1 shows that adults in the poorest quintile are around four times more likely to be problem gamblers than adults in the richest quintile (1.3 per cent compared to 0.4 per cent). Again, there is more of a gradient for FOBT players than for online gamblers; in fact, there is no obvious income gradient for online gamblers, with adults in the richest quintile being just as likely to be problem gamblers as adults in the poorest quintile.

Table 1. Proportion of problem gamblers by household income quintile in England, 2016

Source: Health Survey for England, 2016

Table 2. Proportion of problem gamblers by Index of Multiple Deprivation in England, 2016

Source: Health Survey for England, 2016

4. These findings align with previous research by Landman Economics using the Living Costs and Food Survey (the main source of data on household expenditure by category of goods and services in the UK) which showed that gambling expenditure is a much higher share of household disposable income (3.1 percent) for households in the lowest income quartile than for households in the highest income quartile (0.6 percent). (Reed 2014a, pp8-9)

2               The relationship between location of LBOs and deprivation

5. Research by Landman Economics and Geofutures for the Campaign for Fairer Gambling has found a clear positive relationship between the extent of deprivation in local areas and the number of betting shops in those areas (Reed 2014, pp2-7). Landman Economics has also found similar patterns in the relationship between betting shop location and deprivation for Scotland (Reed, 2016).

3               The wider economic impact of FOBTs

6. Gambling industry representatives and other supporters of the current regulatory framework for FOBTs and online gambling often claim that the gambling industry makes a positive contribution to the economy, and that increased regulation of the sector would have an adverse economic impact. For example, the Association of British Bookmakers has claimed repeatedly that increased regulation of FOBTs would lead to substantial job losses in the betting sector (e.g. ABB 2013). However, this view does not take account of the overall impact of a shift in consumer spending towards FOBTs and away from other goods and services. Each pound which a consumer spends on FOBTs (net of winnings) is by definition a pound which is not spent elsewhere in the economy. In its November 2015 report on the economic impact of FOBTs, Landman Economics estimated the amount of employment supported by a certain quantity of expenditure on FOBTs compared with the employment supported by the same quantity of consumer expenditure on other goods and services in the economy. (Reed 2015, pp10-15).

7. The calculations from the November 2015 Landman Economics report have been updated for this submission using the latest data (details are contained in the Appendix to the submission). Because expenditure on FOBTs supports relatively little employment compared with consumer expenditure elsewhere in the economy, the analysis concluded that £1bn of “average” consumer expenditure supports around 20,000 jobs across the UK as a whole, whereas £1bn of expenditure on FOBTs supports only around 5,000 jobs in the UK gambling sector. This implies that, other things being equal, an increase of £1bn in consumer spending on FOBTs destroys around 15,000 jobs in the UK. The results from the Landman Economics analysis suggest that, if current rates of growth of FOBT expenditure are maintained:

• Gross industry revenues from FOBTs will increase by around 40 per cent over the next ten years, resulting in a gain of around 3,000 jobs for the gambling sector by 2028/29 but a reduction of around 13,000 jobs for the economy as a whole – a net loss of 10,000 jobs.
• At the end of the ten year period, the total annual wage bill in areas where FOBTs are established will be around £360 million lower (in today’s prices) than if FOBT use remained at its 2018 level.

8. At the end of the ten year period net tax receipts will also be around £80 million per year less due to the expansion of FOBTs. Revenue from Machine Games Duty is forecast to increase by around £180 million but this is more than offset by reduced receipts from income tax and National Insurance contributions (due to lower employment) and reduced VAT receipts (due to lower consumer spending on other goods and services).

4               The economic impact of online gambling compared to FOBTs

9. Evidence from statistics published by the Gambling Commission shows that the Gross Gambling Yield[2] (GGY) for FOBTs grew from around £1.3 billion to around £1.9 billion (measured in real terms) between 2008-09 and 2016-17 – an increase of just under 50 per cent in eight years. In 2017-18 the GGY for FOBTs fell to around £1.7bn.  Over the same time period, the online gambling market almost trebled in size in real terms – growing from just under £1.9bn in 2008-09 to just over £5.3bn. The online gambling sector, measured using a definition which includes all online gambling activities, is now over three times the size of the FOBT sector in terms of GGY (Gambling Commission, 2018).

10. Analysis of the relationship between GGY and employment in the online gambling sector compared to betting shops, using Gambling Commission statistics, shows that the amount of employment in the online gambling sector supported per billion pounds of GGY is only around one-tenth that of the amount of employment supported by betting shops. This reflects the relatively low overheads of the online gambling sector – not having to maintain betting shops as a physical presence on the high street, it can generate GGY with far fewer employees.

11. While the 2015 Landman Economics analysis and the updated figures featured in Section 3 above only include the impact of an increase in consumer spending on FOBTs and do not consider online gambling expenditure, it is likely that an increase in expenditure on online gambling destroys even more jobs than the equivalent-sized increase in expenditure on FOBTs. This is because, as explained above, online gambling expenditure supports only around one-tenth the number of jobs that the FOBTs sector does for the same amount of Gross Gambling Yield. This implies that £1bn of expenditure on online gambling supports less than 500 jobs in the UK gambling sector. Using a similar calculation to the one made for FOBTs in the Appendix to this submission, the implication is that an increase of £1bn in consumer spending on online gambling destroys over 20,000 jobs in the UK. Full details of this calculation are contained in the appendix to this submission.

5              Problem and at-risk gamblers and the breakdown of profits from online gambling

12. This section presents estimates of the proportion of profits in the online gambling industry which come from ‘problem’ or ‘at risk’ gamblers. “Problem gambling” is typically defined as gambling to a degree that compromises, disrupts or damages family, personal or recreational pursuits (NatCen, 2018). Gamblers can be classified using the Problem Gambling Severity Index, which is based on gamblers’ responses to a set of nine questions regarding various aspects of problem gambling (NatCen 2018, p65). Each item is assessed on a four-point scale: never, sometimes, most of the time, almost always. Responses to each item are given the following scores: never=0; sometimes=1; most of the time=2; almost always=3. When the scores for each item are summed, a total score ranging from 0 to 27 is possible. Based on the PGSI score, gamblers are classified into four categories:

• Problem gamblers (PGSI score of 8 or more);

• Moderate risk gamblers (PGSI score of 3-7);
• Low risk gamblers (PGSI score of 1 or 2);
• Not at risk gambers (PGSI score of zero).

13. Research by NatCen (2018) uses survey data for England, Scotland and Wales to classify gamblers according to problem/at-risk status using the schema above, and then analyses the extent to which different risk categories of gambler participate in different gambling activities, including online casino gaming and online betting. Landman Economics has combined the data from the NatCen analysis with evidence from a recent PWC report on the on the overall composition of online bets placed by problem gambler status to produce an estimate of the proportion of industry profits which arise from at-risk or problem gamblers[3]. Table 3 below shows the calculations used in producing this estimate. Our overall finding is that 54 percent of the online gambling industry come from problem or at-risk gamblers (with 20 percent being from problem gamblers, 17 per cent from moderate risk gamblers and 17 per cent from low-risk gamblers). Given that only around a third of online gamblers are classified as problem or at-risk according to the data analysed by NatCen, this is a significant finding, and shows the extent to which the online gambling industry is reliant on problem and at-risk gamblers for a majority of its profits. This is another crucial – and worrying – aspect of the economic impact of gambling.

Table 3. Proportion of online gambling industry profits from problem or at-risk gamblers

Notes:

Total annual online gambling expenditure calculated from PWC (2017) Table 13, p43

Percentage of online gambling activity by risk category calculated by Landman Economics using Health Survey for England 2016 data.

Calculation of average profitability uses Gambling Commission (2019) statistics to calculate GGY as a percentage of turnover for online casino and online betting activities. Results show profitability of 3.5% for online casino gaming and 7.9% for online betting activity.

Data from the HSE (2016) indicate that there approximately 2.3 times as many gamblers participating in online betting activity compared to online casino activity.

Response to Question 8

6              Priorities for research on gambling behaviour

14. The fact that Landman Economics had to combine evidence from two different studies – PWC (2017) and NatCen (2018) – to produce the first available estimates (to our knowledge) of the proportion of online gambling industry profits arising from problem and at-risk gamblers, seems to indicate that the focus of current research activity is misplaced. I would recommend that future research focuses more on the economic and social costs arising from problem and at-risk gambling. For example, it would be useful for the Gambling Commission to commission in-depth research on the extent to which the gambling industry – both online and in betting shops and other outlets – relies on problem and at-risk gamblers to drive its business model and provide a large proportion of its profits. This would help supplement and extend the initial findings which I present in Section 5 of this submission.

Summary and conclusions

15. The empirical evidence presented in this submission demonstrates that high rates of problem gambling among FOBT users and online gamblers impose significant economic and social costs on Britain, with the largest impacts being on the poorest and most deprived parts of society. Four findings stand out in particular. Firstly, there are clear links between problem gambling and deprivation and problem gambling and income. Poorer households are more likely to contain problem gamblers than richer households, and households in more deprived areas of England are more likely to contain problem gamblers than households in less deprived areas. The income and deprivation gradients are especially strong for FOBTs. Secondly, there is a clear positive relationship between the extent of deprivation in local areas and the number of LBOs in those areas. Thirdly, there is a negative impact of FOBTs and online gambling on the wider economy through reduced employment and tax receipts because FOBTs and online gambling expenditure do not support as many jobs as most other forms of consumer spending. The estimated negative impact of online gambing is bigger than for FOBTs because the online gambling industry employs far less people per pound of Gross Gambling Yield than Licensed Betting Outlets do. Finally, calculations by Landman Economics based on previously published research by PWC and NatCen suggest that the online gambling industry is reliant on problem and at-risk gamblers for more than half of its total profits.

References

Association of British Bookmakers (2013), The Truth about Betting Shops and Gaming Machines, submission to the DCMS Triennial Review, April 2013

Gambling Commission (2018), Industry Statistics, November 2018.

Gambling Commission (2019), Industry Statistics, May 2019. https://www.gamblingcommission.gov.uk/news-action-and-statistics/Statistics-and-research/Statistics/Industry-statistics.aspx

Ministry of Housing, Communities and Local Government (2015), “English indices of deprivation 2015”. https://www.gov.uk/government/statistics/english-indices-of-deprivation-2015

NatCen (2018), Gambling behaviour in Great Britain in 2016: Evidence from England, Scotland and Wales.

PWC (2017), Remote Gambling Research: Interim Report on Phase II. GambleAware.

Reed H (2014a), Fixed Odds Betting Terminals, Problem Gambling and Deprivation: A Review of Recent Evidence from the ABB. Landman Economics.

Reed H (2014b), “A review of the Local Data Company’s report An independent analysis of betting shops and their relationship to deprivation along with their profile relative to other high street business occupiers, Landman Economics.

Reed H (2015), The economic impact of Fixed Odds Betting Terminals: 2015 Update. Landman Economics.

Reed, H (2016), “The relationship between location of betting shops and deprivation in Scottish local authorities”, Landman Economics.

Responsible Gambling Strategy Board (2017), Advice in relation to the DCMS review of gaming machines and social responsibility measures, https://www.rgsb.org.uk/PDF/Advice-in-relation-to-the-DCMS-review-of-gaming-machines-and-social-responsibility-measures.pdf

Appendix: New calculations of the economic impact of FOBTs compared to other economic activity

16. This Appendix presents new estimates of the overall economic impact of increased expenditure on FOBTs on economic conditions in the localities where the FOBTs are located. The results here are an updated version of Chapter 3 of the Landman Economics report The Economic Impact of Fixed Odds Betting Terminals: 2015 Update (Reed 2015).

A.1 The impact on jobs and economic output

17. The growth in the FOBTs sector over the last two decades has led industry representatives to lobby against greater controls on FOBTs (for example, a reduction in the maximum stake, currently £100 for B2 machines but being reduced to £2 from April 2019) on the grounds that restrictions on FOBTs would reduce growth and lead to job losses in the industry (see for example ABB 2013; ABB 2015).

18. However, it makes no sense, economically speaking, to consider the impact of increased expenditure on FOBTs on the betting sector in isolation from the rest of the economy. Each pound which a consumer spends on FOBTs (net of winnings) is, by definition a pound which is not spent elsewhere in the economy. Hence the question of whether increased expenditure on FOBTs generate increased economic activity or not is really a question about whether each pound spent on FOBTs supports more economic activity than a pound spent elsewhere in the economy.

19. The basic approach taken in this Appendix to calculating the impact of FOBTs on the economy is to estimate the amount of employment supported by a certain quantity of consumer expenditure on FOBTs compared with the employment supported by the same quantity of consumer expenditure on a weighted basket of other goods and other services in the economy. Thus, rather than asking the question “how much economic activity is created by Fixed Odds Betting Terminals?” the analysis here asks, “what is the change in economic activity if consumer expenditure shifts from other goods and services to FOBTs?” In terms of the aggregate economic impacts of FOBTs on the UK economy, the latter question is much more appropriate than the former.

20. Note that the focus here is explicitly on local economies; the analysis draws a distinction between expenditure on wages, which (if betting shop employees live reasonably locally) is likely to be “re-circulated” into the local economy via consumers spending a proportion of what they earn, and profits for the betting industry, which (given that most betting shops are owned by large-scale national chains) are not likely to be re-spent in the local economy.
21. The analysis in this Appendix proceeds by attempting to calculate what proportion of Gross Value Added (GVA - a measure of economic output used by the UK Office for National Statistics – essentially equal to net industry revenue after subtracting costs of production) from FOBTs is accounted for by wage costs. This “share of wages in GVA for FOBTs”  is compared with the proportion of GVA from consumer expenditure in the UK economy as a whole which is accounted for by wage costs (the “share of wages in GVA for overall consumer expenditure”). To the extent that £1 of expenditure on FOBTs supports fewer jobs than the “average” £1 of consumer expenditure, an increase in spending on FOBTs will reduce overall employment and economic activity.

22. The following assumptions are made about the amount of employment supported by Fixed Odds Betting Terminals:

• It is assumed that each set of 4 FOBTs supports one full-time job at the average hourly wage rate for people working in the gambling industry on an hourly rate (rather than a salaried basis). According to the 2017 Annual Survey of Hours and Earnings[4], the median annual salary for full-time employees in the gambling industry is £21,100. In practice is likely that FOBTs, as a completely automated gambling format, support less employment than this, but given that there is a maximum of 4 FOBTs per betting shop it seemed reasonable to apportion at least some cleaning and maintenance time for each shop to maintain the FOBTs and the environment around them, as well as allowing for some of the tasks undertaken by counter staff in betting shops to support FOBT play (e.g. use of debit cards rather than cash to fund play, “selling” the machines to customers by offering free play sessions and tournaments as marketing tools, and so on.)[5]
• Gross Value Added from FOBTs is estimated by using a statistic for GVA for the entire gambling industry (including bookmaking, casinos, betting and online gambling) from the DCMS publication DCMS Sector Economic Estimates (DCMS, 2018b) and apportioning GVA in line with the share of total gross revenue from FOBTs as a share of total revenue for LBOs[6].

23.         Table A.1 shows the calculation of the share of wages in Gross Value Added for the Fixed Odds Betting Terminals industry and compares this with the share of wages in GVA across UK private sector industries excluding financial services[7].

Table A.1. Share of wages in Gross Value Added for FOBTs compared with average across UK private sector industries

Notes:

FOBTs GVA calculation based on DCMS’s estimate of GVA for the gambling sector in 2017 (DCMS 2018b), allocated pro-rata to FOBTs on the basis of data from Gambling Commission (2018) showing that betting shop activities (including OTC betting and FOBTs but excluding online betting) account for approximately 28% of total gross revenue for the gambling industry, while FOBTs account for 58% of gross revenue from betting shops. Employment costs for FOBTs calculated assuming one full-time employee per set of 4 FOBTs at annual wage of £21,100.

Entire UK private sector GVA and employment costs calculations calculation based on data from 2017 ONS Annual Business Survey data for SIC2007 industries B (mining), C (manufacturing), D (electricity and gas), E (water), F (construction), G (wholesale and retail trade), H (transport and storage), I (accommodation and food services), J (information and communication), K (finance and insurance), L (real estate), M (professional scientific and technical activities), N (administration), R (arts and entertainment) and S (other service activities) summed together. Industry K (finance and insurance) is currently excluded from the ABS dataset due to concerns regarding data quality.

24. Table A.1 shows that the total share of wages in Gross Value Added for Fixed Odds Betting Terminals, under our assumptions, is 10.5 percent – much lower than the share of wages in Gross Value Added for the UK private sector (excluding financial services) overall, which is approximately 49 percent. The implication of these figures is that consumer expenditure on FOBTs supports very little employment compared with an average basket of consumer spending on goods and services. If one pound of consumer spending is diverted from other goods and services to FOBTs, it is likely to support only just over one-fifth as much employment as it would have done, on average, if that pound had been used to buy other goods and services. The corollary of this finding is that FOBTs deliver particularly high profits for bookmaking firms because wage costs required to support FOBTs are so low relative to the amount of revenue that they generate.

25. In terms of overall employment generation, what is the impact on local economies of a shift of consumer spending into FOBTs? Taking into account average wages in the betting sector compared to average wages across the UK private sector, this analysis finds that £1bn of “average” consumer expenditure supports around 20,000 jobs across the UK as a whole, whereas £1bn of expenditure on FOBTs supports only around 5,000 jobs in the UK betting sector. This implies that, other things being equal, an increase of £1bn in consumer spending on FOBTs destroys just over 15,000 jobs in the UK. Furthermore, the jobs created in the UK betting sector are on average lower paid (average full-time annual salary around £21,000) than jobs created by consumer expenditure on other goods and services (average full-time annual salary around £37,000[8]).

26. This is important in terms of the likely expansion of the FOBTs industry over the next decade, if rules governing maximum stakes stay as they currently are. Table A.2 extrapolates the trend in Gross Gambling Yield from the period 2008/09 to 2014/15 to provide estimates of total gambling yield from FOBTs in 2015/16 (the current financial year) and 2025/26 (ten years from now). The Table shows the implied growth in GGY from 2015/16 onwards, and the implied loss of jobs across the economy as a whole resulting from this expansion of FOBTs in the betting sector.

Table A.2. Implied growth in FOBTs business and economic impact at current rates of growth

Notes: all figures at April 2018 prices.

Source: author’s own calculations

27. Table B.2 suggests that Gross Gambling Yield from FOBTs will increase from £1.7 billion to £2.4 billion over the next ten years, resulting in a gain of around 3,000 jobs in the betting sector but a loss of around 13,000 jobs elsewhere in the economy, leading to an overall net reduction of around 10,000 jobs for the economy as a whole by 2028/29.

28. Over the ten year period, the impact of the expansion of FOBTs in terms of reduced wage payments to people working in the local economies where FOBTs are established is to reduce the total wage bill in these areas by around £400 million by 2028/29. This is due to a combination of two factors: (a) the reduction in the total number of jobs supported by consumer spending as a result of switching spending from other goods and services into FOBTs, and (b) the fact that jobs arising as a result of the expansion of FOBTs are relatively low-wage compared with jobs supported by other types of consumer spending.

A.2              Impact of increased FOBTs on tax receipts

29. One important aspect of the economic impact of increased numbers of FOBTs is their impact on tax revenues. This report models three main revenue impacts of a shift in consumer expenditure towards FOBTs:

(1) Increased receipts of Machine Games Duty (MGD) – this is paid at a rate of 25% on gross revenues from category B2 gaming machines (following an increase from the previous rate of 20% in March 2015).

(2) Reductions in VAT receipts arising from reduced consumption on goods and services elsewhere in the economy, the majority of which attracts VAT at the standard rate of 20%[9].

(3) Reductions in income tax and National Insurance Contributions (NICs) arising from reduced overall employment in the UK economy (as explained above), meaning that there are fewer people in work to pay income tax and NICs to the UK Exchequer.

30. Table A.3 adds these tax revenue impacts together to calculate the total impact of the  expansion of FOBTs on tax revenue over a 10-year period (up to 2028/29).

Table A.3. Impact of increase in FOBTs on per-year tax revenues over a 10-year period

Notes: Machine Games Duty revenues calculated as 25% of the increase in Gross Gambling Yield over 10 years using GGY figures in Table B.2.

Income tax and NICs revenues calculated assuming that the average full-time weekly wage of additional workers taken on in the betting sector is £21,100 per year, whereas the average wage of workers made redundant in other sectors of the economy is £36,600 per year.

Reduced VAT revenue calculated on the basis that 52 percent of consumer expenditure shifted from other goods and services to VAT would have attracted VAT at the standard rate of 20% (House of Commons Library, 2012).

31. Table A.3 shows that although the expansion in FOBTs over the next decade is estimated to lead to increased MGD revenue of around £180m, this is accompanied by a reduction in income tax and NICs revenue of around £150m and reduced VAT revenue of around £105m, meaning that total tax revenue decreases by just under £80m.

6 September 2019