Productivity Growth and the Funding of Public Service Broadcasting in the UK

1. Productivity Growth and the Funding of Public Service Broadcasting in the UK David Paton Nottingham University Business School Leighton Vaughan Williams Nottingham Trent University May 2008

2. Introduction • The role of PSB and how it should be funded has long been a source of controversy amongst policy makers in many countries. • Typically, the public service broadcaster has an incentive to argue that improvements in, for example, programming quality should come from increases in public subsidy whereas public finance officials would prefer to see improvements funded from efficiency savings.

3. Introduction (cont.) • The negotiating process is complicated by the difficulty in identifying a level of efficiency savings which it is reasonable to expect the public service broadcaster to achieve.

4. Why Public Service Broadcasting? • Classic Market Failure argument

5. Market Failure • 1. Non-exclusivity (although the technology exists to exclude and is used by some commercial broadcasters). • 2. Non-rivalry (i.e. consumption of the good by one person does not reduce its availability to others). • The ‘Free Rider’ problem, in which those who do not pay are as able to consume the product as those who do pay.

6. Market Failure (cont.) • ‘Merit good’ principle, i.e. a good whose value to an individual (in terms of information and education, for example) exceeds the value placed on it by the individual, in part because people are not fully informed. • In a more general sense, PSB may provide more positive externalities (e.g. improvements in social responsibility) than would a free market which might pander to the lowest common denominator.

7. Rationale (cont.) • The rationale for PSB may become even more important as spectrum width and digital technology reduces barriers to entry, although advantages of scale cold mean that a high level of market concentration is maintained.

8. Citizen-Based Grounds for PSB • All this is part of what has been termed (Cave, 2004) a “citizen-based” ground for public service broadcasting. • Notably, the public interest objective of increasing overall levels of programme quality, creating a “… bias towards quality” (Noam, 1987).

9. Department for Culture, Media and Sport directive • PSB should “… provide a strong and distinctive schedule of benchmark quality programmes on all its services (DCMS, 2000).

10. Why study BBC? • 1. The UK is unusual in funding its main public service broadcaster, the BBC, by means of a licence fee levied on anyone (with defined exceptions) who owns a TV set. • 2. The BBC is the largest and most dominant public service broadcaster in the world.

11. Background • 1. Founded in 1922 as the ‘British Broadcasting Company’ by a group of wireless manufacturers. • 2. Royal Charter granted in 1927. • 3. TV broadcasting commenced in 1936. • 4. Charter most recently renewed in 2006.

12. Funding Licence fee paid for by every household with TV, with defined exemptions, e.g. for those over 75. BBC income, 2006-07 = £3.24 bn (BBC Executive Report, 2007). £3.1 bn (2005-06); £2.94 bn (2004-05). The UK both devotes a larger share of GNP to PSB, and attracts a larger audience share, than any other developed economy.

13. Level of funding • Determined every 5 years by a ‘licence fee settlement’ resulting from negotiations between central Govt and the BBC. In these negotiations the Govt has increasingly been concerned with trying to assess the level of efficiency savings which the BBC might reasonably be expected to achieve over the lifetime of the ‘settlement’.

14. Measurement Problems • Given that the BBC’s programming output is free at the point of use, it is difficult to identify levels of output in a form suitable for assessing productivity levels or growth in the BBC. • A natural alternative is to estimate productivity growth in the commercial broadcasting sector and to use these as a benchmark to judge reasonable efficiency gains which the Govt might expect the BBC to make. • Even so, there remains a problem in defining measurable units of output and adjusting for quality changes.

15. Motivation: Summary • Funding settlement for public service broadcasting (PSB) in the UK subject to political tensions. • Broadcasters want more public funds whilst Treasury want efficiency gains • Treasury face difficulties in identifying what efficiency gains are reasonable. • We are interested in estimating productivity growth rates amongst commercial broadcasters to use as a benchmark for reasonable PSB efficiency savings.

16. Previous work • 1. Very few estimates of productivity growth in broadcasting and, to our knowledge, none at all based in the UK. • 2. Exceptions include Triplett and Bosworth (2003) who calculate labour productivity in a range of US service industries. They use data from the Bureau of Economic Affairs (BEA) to calculate an annual growth rate of labour productivity in radio and TV broadcasting of 1.2% p.a. between 1995 and 2000. Using Bureau of Labor Statistics (BLS) data, they report a similar growth rate , of 1% p.a.

17. Previous work (cont.) • Another exception is Sichel (2001) who report labour productivity grwoth estimates using both total output and value added for th broadcasting sectorfor three periods: • Total output: 1977-90 (0.6%); 1990-95 (1.1%); 1995-99 (0.7%). • Value added: 1977-90 (-0.8%); 1990-95 (6.3%); 1995-99 (-4.5%).

18. Summary of previous work • With the exception of Asai (2005), who looks at broadcasting productivity in Japan, we have no estimates of broadcasting productivity for any period beyond the year 2000 or of TFP and no estimates at all for broadcasting in any country except the US. This is particularly important in the context of this study given the institutional differences between broadcasting in the US and UK.

19. Methodology 1. Labour productivity growth estimates: compare broadcasting with broader sectors: “Other services” and “Recreational, Cultural & Sporting Activities” 2. TFP estimates using Stochastic Frontier Analysis (SFA) method, decomposed into technical change and efficiency change. 3. TFP also estimated using Levinsohn-Petrin technique (see Levinsohn and Petrin, 2003, for details).

20. Data Annual Respondents Database (ARD) – a plant-level file based on the Annual Business Inquiry, a survey conducted by the OFS. Firms selected for inclusion in the ABI from the IDBR at the ONS. Sampling is based on size by employment on the Register. Sampling undertaken at “reporting unit”… … reporting unit selected by enterprise. Broadcasting data available from 1997 to 2003 for around 130 firms each year.

21. Measurement of Variables Output measures: Gross Output (GO) Gross Value Added (GVA) (see paper for construction)Inputs measures: Employment Capital stock Materials (for GO only)Deflate by CPI for Recreation & Leisure (published by the ONS)

23. Results 1. Labour productivity growth estimates 2. SFA TFP growth estimates & decomposition 3. Comparison using alternative estimators (e.g. Levinsohn-Petrin) 4. TFP estimates by employment group 5. Alternative price deflators (not reported here)

24. 1. Labour Productivity Growth estimates Notes: (i) Figures are mean annual % growth for the specified periods

25. Summary of Results 1. Mean annual labour productivity growth over 1997-2004 estimated to be 4.8% using GO and 6.0% using GVA.

26. 2. SFA TFP growth estimates & decomposition Notes: (i) Figures are annual % growth for the specified periods

27. Summary of Results (cont.) 1. Mean annual labour productivity growth over 1997-2004 estimated to be 4.8% using GO and 6.0% using GVA. 2. Mean TFP growth rate using SFA is 11.9% (GO) and 5.4% (GVA). Growth explained mainly by technical change but also by efficiency improvements

28. 3. Comparison using alternative estimators Notes: (i) Figures are mean annual % growth for the specified periods (ii) Levinsohn-Petrin (LP) estimates allow for endogeneity of inputs. (iii) GO estimates and the L-P GO and GVA estimates are based on 1999-2000, 2001-2003 and 1999-2003 respectively.

29. Summary of Results (cont.) 1. Mean annual labour productivity growth over 1997-2004 estimated to be 4.8% using GO and 6.0% using GVA. 2. Mean TFP growth rate using SFA is 11.9% (GO) and 5.4% (GVA). Growth explained mainly by technical change but also by efficiency improvements 3. Levinsohn-Petrin estimates broadly comparable (although implausible for TV alone).

30. 4. TFP estimates by employment group Notes: (i) Figures are mean annual % growth. (ii) GO estimates are based on 1999-2000, 2001-2003 and 1999-2003 respectively.

31. Summary of Results (cont.) 1. Mean annual labour productivity growth over 1997-2004 estimated to be 4.8% using GO and 6.0% using GVA. 2. Mean TFP growth rate using SFA is 11.9% (GO) and 5.4% (GVA). Growth explained mainly by technical change but also by efficiency improvements 3. Levinsohn-Petrin estimates broadly comparable (although implausible for TV alone). 4. TFP estimates reasonably stable across firm sizes

32. Policy Implications • Broadcasting sector in UK has experienced positive productivity growth over recent years • Technical change and (less so) efficiency catch-up contributed to productivity growth. • Likely to be potential for significant efficiency savings from the BBC • Lower licence fee increases !! • But caution needed due to relatively small sample size & rapid changes to structure of industry.