R

1
RD and productivity in the UK Evidence from
firm-level data in the 1990s

• Mark Rogers,
• Harris Manchester College, Oxford
• and Oxford Intellectual Property Research Centre
• Full paper and slides at
• http//users.ox.ac.uk/manc0346/
• mark.rogers_at_hmc.ox.ac.uk
• ONS Conference Septmebr 2005

2
UK Business RD / GDP
UK business RD (BERD) relatively low with
respect to G5. Slow decline in 1990s (average
1.30) 1980s (1.47)
3
Further issues in BERD fall

• Defence-related BERD has fallen in 80s/90s, but
  civil BERD also small decline
• Industry structure does not account for
  differences with G5 UK firms have lower
  intensity as well. UK SME sector has relatively
  high BERD larger firms reverse.
• Real RD spend increased in pharma. (1987 was 10
  of total, now around 25). Services also
  increasing.
• UK has high of foreign funded BERD (22 in
  1999 Germany 2.7)

These statements based on government analysis
in HM Treasury Science and Innovation Framework
2004-14. No formal, analytical papers.
4
This paper

• Data on large UK companies (1989-2000) (max
  n719, balanced and unbalance panel)
• These data 10.7bn RD in 2000, ONS estimate
  11.5bn BERD for entire UK.
• Analyse (gross marginal) rate of return to RD
• Why interested?
• Bond et al (2002) find that UK rate of return is
  higher than Germany
• implies constrained RD (financial or human
  capital) ? specific policies
• This consistent with older, anecdotal view,
  that City of London not interested in RD.

5
Evidence on rates of return

• US studies (to mid80s) 0.13 to 0.25
• German/French studies 0.22 to 0.35
• Based on production function approach

6
from Table 4 in paper
7
UK rate of return?

• Above shows 0.18 to 0.25
• Have run IV and GMM models (RD coefficients 0.07
  to 0.25 labour and capital coefficients
  volatile)
• Omitting highest RD intensity firms does raise
  coefficient (to around 0.30)
• Conclusion, no strong evidence that UK firms have
  higher rates of return than German/French

8
Did rate of return change in 1990s?

• Firm-level data used here show temporary fall in
  RD/value added in 96/97 do not mimic aggregate
  data.
• Using dummy variables for years, and interacting
  with RD, find no evidence of changes in returns
  over 1990s
• Overall, no evidence of shifts in constraints on,
  or demand for, RD during the 1990s

9
Financial constraints?

• Firms that are credit constrained in RD, should
  have higher rates of return (i.e. profitable RD
  projects they cannot fund)
• Bond et al (2003) find no constraints for firms
  that do RD, but low past cash flow implies less
  likelihood of doing RD
• Analyse difference between stock market listed
  and non-listed, and also past profitability (NPBT
  t-2, t-3)
• Listed status has no significant role.

10
Past profitability and returns to RD
Coefficient on RD/value added stable No
evidence that RDers are credit constrained
11
Competition and returns to RD

• Basic theory suggests returns should vary with
  competitive conditions
• Very difficult to measure competition
• Greenhalgh and Rogers (2004) use profit
  persistence and Pavitt sector methodology
• Profit persistence uses pit fi bpit eit
• low b, means rapid adjustment back to normal
  profits, high competition
• Analyse this for six Pavitt sectors

12
Pavitt sectors
13
Low ? high competition
Low return in (competitive) science-based
sector Pharma. firms in sector 4 have returns of
0.14 UK science-based sector is unusual
14
Conclusions

• Results
• UK returns to RD roughly same as other
  countries.
• No evidence of changing returns in 1990s
• Listed status or past profitability no influence
  on rates of return
• Science-based sector has lowest returns, and is
  also most competitive (according to profit
  persistence)

15
Implications

• Why is UK BERD low?
• No evidence of high returns, implies finance/HK
  constraints not the problem
• Fact that returns similar to other countries, yet
  RD low, implies low appropriability weak
  demand
• Excessive competition? (but intl markets?)
• Could be poor management and weak use of IP
• Further work
• role of international firms
• role of IP
• further analysis of pharmaceuticals, electronics
  and electrical

16
Profit persistence analysis
Use PBT/Sales as p
High competition b ? 0 Low competition b ? 1