Written evidence submitted by David Connell (BER0006)

Building a UK Mittelstand; Increasing the Role of Procurement and other Demand Led Innovation Measures in the Overall Policy Mix

David Connell, Senior Research Fellow, Centre for Business Research, Cambridge Judge Business School. Author of Leveraging Procurement to Grown the Innovation Economy, November 2017, final report of an independent review of SBRI commissioned by 10 Downing Street. This submission draws on my 30 years’ business experience with STEM based companies of all sizes and as a venture capital investor, together with research and publications on start-up and growth strategies, strategic partnering, intermediate research organisations, the commercialisation of academic research, and innovation policies in the UK, US, Germany, and Asia, mainly undertaken with the Centre for Business Research, Cambridge Judge Business School.

1. Introduction and Summary

The Government’s 2.4% R&D target only has value to the extent that it leads to the growth of profitable STEM based companies delivering significant economic value to the UK over the long term, through employment and taxes, and with high levels of exports and self-funded R&D.

Given the decline of UK manufacturing, we need to create a new cohort of UK Mittelstand companies to begin to achieve this and rebalance the economy.

This submission focuses on how the UK’s most economically significant new STEM based companies have started, grown and been financed. It considers what policy lessons can be learned from these success models, and it examines the relevance of current innovation policies to them. It argues that UK innovation policy needs to evolve to support these success models better if we are to create more such Mittelstand companies.

To do so we must not just address weaknesses in individual programmes, but also examine critically the cost effectiveness of overall innovation spending across government, including BEIS, Treasury and spending departments. This is a task which successive governments have failed to undertake.

The submission concludes with a series of costed proposals to increase the cost effectiveness of total UK innovation policy spend, in particular, by giving more emphasis to increasing the demand for innovation by encouraging customer funding. To afford this we need to modify and/or downscale poorly performing policies designed to increase the supply of innovation through subsidies aimed at making it cheaper to undertake, or through policies that are excessively dependent on technology push from university research.

Supporting evidence for the analysis is cited in the appendix.

2. Start Up Models, Funding Strategies and Geography

STEM based companies can start in different ways. Most policy and media attention is focused on the hard start-up or Silicon Valley model, in which venture capital is raised to start the business based on a well-defined plan to commercialise an invention, and further rounds are raised until the business can be sold or floated on a stock exchange.

At the other extreme is the soft start up model in which, rather than raising venture capital, the team negotiates a series of progressively more ambitious consulting and R&D contracts based around their expertise. Over time, they may then develop product businesses, often by building on one-off contracts for individual lead customers.

In between are various hybrid models, with venture capital being raised to fund the transition to products or the creation of product oriented spin-offs, and the use of lead customer contracts to customise a standard product for early applications.

The problem with the hard start up model is that successful companies are nearly always sold. And the more significant the technology, the more likely the sale will be for its strategic value to a large corporation, before the UK start-up has grown significant revenues to build a complete business employing teams with a range of operational and support skills around its R&D. Inevitably, most such sales will be to corporations headquartered abroad, leading to the truncation of entrepreneurial growth within the UK.

In this respect venture capital has a different economic impact in the UK compared with the US, where acquisitions of early stage companies are more likely to retain their future growth potential within the US.

This is an inevitable consequence of the UK’s position in the world economy[1] and the need for venture capital funds to make the kind of home-run multiples needed to deliver an acceptable return for their own investors. The rapid rate of technological change means there may only be a small window in which such exits can be achieved.

In contrast, entrepreneurs pursuing a soft start up strategy are more likely to be able to delay, minimise or avoid venture capital, enabling them to retain control and, should they wish to, go on to build a substantial UK business rather than selling out early.

Whether a company is venture capital backed or not, a lead customer development contract helps focus R&D on real customer needs. And it brings both money and endorsement for future customers, partners and investors, accelerating growth and reducing both the breadth and depth of the valley of death. It is the kind of transformational event all entrepreneurs and investors look for.

Research on the Cambridge Cluster shows that soft start-ups and lead customer contracts have played a key role in a wide range of sectors, from semiconductors to life sciences.[2]

The histories of some of the UK’s most economically significant new STEM based companies to have been started in recent decades[3] show that:

• Lead customers, from both the private and public sectors have, in different ways, played a key role in originating and funding the development of their products;
• Many started in specialised or unfashionable fields where there was little venture capital or government funding available;
• Many either raised no venture capital investment or delayed it until growth was underway; in two cases founders and their families still have controlling shareholdings;
• Growth has taken place over a long period rather than in spectacular bursts.

Although some of these firms have recently been acquired, this is after many years of operating as successful, full-function businesses.

Other examples are legion, including:

• Vodafone, the UK’s most successful start-up since the second world war, was a spin out from Racal, which began as a specialist two man R&D company developing radio technology for the MOD;

• Microsoft, whose first product, an operating system for the IBM PC, was developed under contract, and then sold to clone manufacturers;

• Intel, whose success was based on the world’s first single chip processor developed under contract for a Japanese calculator company.

It is essential that government innovation policy does all it can to encourage this sort of demand led innovation funding amongst both public and private sector customers.

3.  Summary of Key Innovation Policy Challenges

Increasing UK R&D intensity from the current level of 1.7% to meet the Government’s 2.4% target cannot be achieved without a major rethink of both current policies and the assumptions underlying them. 

•         2.4% target only achievable if government spending leverages much more private R&D spending per £ than at present;

•         Policy must deliver the creation and growth of more complete businesses, not just R&D centres, in particular a new cohort of UK STEM based, Mittelstand firms;

•         Geography means venture capital cannot deliver the same impact on the UK economy as in the US and UK science is likely to be commercialised elsewhere;

•         Non-dilutive finance for start-ups is needed to help prevent successful entrepreneurs being forced by financial investors into early trade sales, with the resulting truncation of growth in UK based operations (recent “Patient Capital” measures do not appear to address this);

•         Policies must encourage the soft start-up model and use of lead customer technology development contracts (from both public and private sectors) on which a high proportion of UK successes have been based;

•         A more integrated approach is needed across Treasury and BEIS policies to ensure the cost effectiveness of overall spending is maximised (the poor economic impact of R&D tax credits is of particular concern);

•         A shift is required in the balance of policy expenditure to increase the demand for innovative technologies as opposed to its supply (See Exhibit 1);

•         Implementation of an effective policy is needed to realise the potential of public sector lead customers through procurement. After 18 years of attempts, a new approach is needed;

•         Individual policy instruments must be redesigned to maximise impact.

4. UK Government R&D Spending; The Big Picture

Between 1986/7 and 2012/13, the Research Councils increased their share of total government R&D spending from 28% to 58%, mainly at the expense of spending departments like Defence, rather than the DTI and its successors. The percentage of total government R&D expenditure going to companies more than halved from around 42% to around 18% over the same period.[4]

This submission focuses on government funding for business R&D through defined innovation policy instruments.

Exhibit 1 Breakdown of Annual Government Funding for Business R&D by Principal Policies (2015/16)

Source; Leveraging Procurement to Grow the Innovation Economy, David Connell. BEIS November 2017 (Innovate UK data 2015/16; HMRC R&D Tax Credit data updated for FY 2015/16)

5. R&D Tax Credits

The cost of R&D tax credits was £2.9 billion in 2015/16. The multiple improvements in the coverage and generosity of R&D tax credits made since they were introduced in 2000 have failed to increase the UK’s business R&D intensity from its starting figure of around 1.1%. Instead the government subsidy seems merely to have substituted for aggregate spending from company generated funds. In 2015/16 they paid for 14% of total business R&D.[5]

Exhibit 2 Impact of R&D Tax Credits on “Net” Business R&D Spending as a Percentage of GDP

Source; HMRC R&D Tax Credit Statistics

To understand the possible reasons behind the poor performance of R&D tax credits it is useful to explore their likely impact on different kinds of beneficiary.

5.1  Impact of R&D Tax Credits on Large Corporations

•         R&D tax credits are handled by group finance functions;

•         Group managements responsible for budget allocations between ongoing operations, dividends and growth;

•         Growth funding split between business development, investment in physical assets, business acquisition and technological innovation;

•         Technological innovation can be via in-house R&D, licensing, or acquisition of companies based in any geography; uncertainties over higher risk R&D make risk adjusted DCFs meaningless so R&D investment is inelastic with respect to cost;

•         Minimum market size tests (typically a billion dollars) and other factors make new ventures unattractive; later acquisition is usually a better option.

5.2  Impact of R&D Tax Credits on Venture Backed R&D Intensive SMEs

•         R&D spending high, so tax credits a useful subsidy for investors;

•         Should help to offset impact of unacceptably low average UK venture capital returns that have persisted for at least 30 years (and which is itself due to the dynamic effect on global competitiveness of start-ups based in small markets like the UK c.f. the US);[6]

•         Effect on institutional and fund of fund investors’ asset allocations depends on long term average asset class returns so any increase in funding for UK VC is unlikely to be rapid at best;

•         Probably a better way of encouraging early stage UK venture capital than government investments in funds of funds on standard terms;[7]

•         BUT technologically successful, early stage companies dependent on venture capital likely to be sold early to foreign corporations with truncation of further growth potential in UK. Lead customer and other non-dilutive early stage funding measures likely to be more cost effective in this respect.

5.3  Impact of R&D Tax Credits on SMEs That Have Not Raised Venture Capital

•         Includes start-ups “funded” through sweat equity, small amounts of angel funding or pursuing a soft start up model based on consulting and technology development contracts AND many larger SMEs with profitable businesses based on services, but with a desire to develop standard products on the back of lead customer contracts;

•         Spending on R&D too small to attract R&D tax credits of sufficient value to have a significant impact;

•         The origins of many of the UK’s most economically significant new Mittelstand companies;

•         Raising venture capital may not be a realistic or sensible way forward for these companies;

5.4  Impact of R&D Tax Credits on Inward Investors and Geographically Mobile R&D

•         Availability of high quality scientists and engineers probably the main driver of location decisions;

•         UK is competitive wrt other cost factors including salaries[8], exchange rates, corporate tax rates and other government incentives including the Patent Box (cost £652m in 2014/15, up from $365m in 2013/14);

•         Key industrial competitors – Germany, Sweden, Finland, Switzerland – have no R&D tax credits. US programme based largely on growth in R&D spending (c.f. absolute spending for UK scheme) and much less generous overall;

•         50% of UK business R&D already carried out by overseas headquartered companies. This is arguably too large.

6.   BEIS/Innovate UK Grants for Businesses

6.1 Recent Positive Developments

•         Increased focus, and relevance of themes through better sector and technology knowledge;

•         Possibly larger grants and increased levels of subsidy (data unavailable);

•         Increased emphasis on business, rather than academically led projects (though reality is hard to assess), and other improvements;

•         Linkages with Catapults and Independent Research Organisations;

•         An imaginative range of experimental programmes linking grants to third party venture capital investment and Innovate UK loans.

6.2 Weaknesses

•         Need for match funding can make grants largely irrelevant to SMEs with a soft start-up or consulting model;

•         Continued preference for collaborative R&D grant model (CR&D)

• The need for at least two participants can lead to artificial collaborations that take time to put together, with each “collaborator” largely pursuing their own agenda and timescale;
• Funding participating organisations in parallel is not conducive to effective project management. A prime contractor/subcontractor model is normal for businesses (and for DARPA);
• CR&D more appropriate for industrial research, rather than close to market demonstrators, prototypes and customer trials; weak commercialisation championship likely;
• Implicit (and incorrect) assumption that innovation in successful start-ups and larger businesses is largely driven by academic research;
• Grant model does not employ phasing with clear go/no go milestones (c.f. normal business approach and SBRI);

•         Current and planned future use of single company grants, like the well regarded, discontinued SMART grant family, is unclear;

•         Long term outcome monitoring using verifiable measures is poor;

•         Past entrepreneurial UK successes have often been in specialised or unfashionable areas. Possible risk that BEIS Challenge areas are passé and in ultra competitive markets. Lead customers likely to be better at picking emerging opportunities for UK businesses than government.

7.    Procurement/SBRI

•         Procurement based funding of R&D is the dominant US innovation policy;

•         Repeated calls for similar policies in UK over two decades including from all major political parties;

•         The (misnamed) SBRI is de facto the only UK innovation procurement policy;

•         Five relaunches since 2000 aimed at meaningful adoption across the public sector;

•         My Review of SBRI shows that, where operated in accordance with SBRI funding guidelines, it can be very effective for both companies and public sector funders.[9] It is frequently transformative for start-ups;

•         As it provides 100% funding in meaningful amounts (typically £1m) SBRI can act like seed capital, including triggering start-ups. (US SBIR, on which SBRI is based, is often described by the SBA as “America’s Seed Fund”);

• Most spending departments have only participated in SBRI intermittently, under temporary pressure from Downing Street. Budgets and individual contracts are frequently too small to be effective. Weak departmental ownership and management. Innovate UK now has the largest SBRI spending level, although it is not a potential customer for the resulting technologies;

•         Continuing calls for expansion of SBRI from Select Committees and industry associations;

•         Spending departments do not regard it as their job to fund innovation in businesses and BEIS does not regard it as its job to fund innovation for spending departments; only the Treasury can cut this Gordian knot.

8        Catapults, Universities and Regional Economic Development

•         Some evidence that Catapult type intermediate research organisations and technology consultancies can be important catalysts of longer term cluster development; [10]

•         Some Catapults failing to implement the demand pull element of the Fraunhofer model, on which they are based, by undertaking customer funded technology development contracts;[11]

•         Need to encourage commercial market for consultancy and contracts by university STEM academics as a stepping stone towards entrepreneurial start-ups and new Catapults in different parts of the UK.

9   Proposals

The aim of the proposals outlined below is to:

•         Strengthen the demand for innovation by both public and private sectors, especially through lead customer technology development contracts;

•         Increase non-dilutive funding for start-ups to reduce later investor pressure for early technology sales;

•         Increase commercialisation and career options for academic scientists by boosting the technology development consulting market (a middle way);

•         Help Catapult’s develop the demand pull side of the model;

•         Increase additionality of R&D tax credits through focus on higher risk R&D;

•         Achieve this with small or neutral overall spending impact, through savings in future government funding of Catapults and research universities, plus efficiency savings from public sector procurements of innovative technology.[12]

9.1  R&D Tax Credits

•         Reform to encourage additional R&D activities (particularly higher risk R&D) rather than subsidising ongoing, more routine R&D;

•         Pay companies (say) 20% of value of R&D tax credits received above (say) £3-500k a year in the form of vouchers. Aim for £300m per annum in vouchers;

•         Vouchers to be used, at choice of recipient, to pay for lead customer, technology development contracts with:

• universities
• independent research organisations, including Catapults
• SMEs;

•         “Lead customer” definition to include IP rights enabling further commercialisation of technology platforms by contractor.

9.2  Non-Dilutive Grant Funding for Single Companies

•         Introduce similar competitive grants programme to well regarded EU SME Instrument [13] (say £50m per annum):

• £50-100k for Phase 1 at 100% support rate
• Up to £2m for Phase 2 at 70% support rate
• Interviews at Phases 1 and 2
• Close monitoring similar to NHS SBRI;

•         Introduce single company competitive grant programme with grant payments match funded with payments by corporate lead customers (say £50m per annum)

• Use of R&D tax credit vouchers allowed
• This has similarities with the new Innovate UK grant programmes with match funding from equity investors and government loans.

9.3 Procurement: Implement Recommendations of Independent Connell Review of SBRI in Full

•         New central fund growing to £200m per annum into which departments and agencies can bid to run multi-year programmes. Held by Cabinet Office to give leverage over spending departments and ensure ownership;

•         Small fund management board, including Innovate UK and independents with a business and VC background to ensure effective, embedded, departmental programmes using best practice from the private sector;

•         Phasing of contracts and interviews to focus funding on best projects;

•         Adherence to guidelines, including 2 year Phase 2s with enough to make a difference (typically £250k to £1m). New Phase 3; the devil is in the detail;

•         Use across public sector, including less obvious agencies where the UK has world class lead customers e.g. BBC, academic research community (NB ad hoc contracts from these have historically led to UK business successes).

September 2018

APPENDIX: SUPPORTING ANALYSIS AND EVIDENCE

1. Leveraging Procurement to Grow the Innovation Economy; An Independent Review of the Small Business Research Initiative; David Connell.  November 2017. BEIS

123 pages; covers economic rationale, relationship with other UK innovation and procurement policies, US SBIR and use of SBRI across different departments and agencies and impacts. Case histories of UK and US programme beneficiaries

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/669605/Leveraging_Public_Procurement_David_Connell_report.pdf

2. Creating Markets for Things that Don’t Exist; The Truth About UK Government R&D and How the Success of SBRI Points the Way to a New Innovation Policy to Help Bridge the Valley of Death and Rebalance the UK Economy; David Connell. November 2014. BEIS. Centre for Business Research, Cambridge Judge Business School.

68 pages. Chapter 5 analyses the “triple whammy” long term decline in UK government spending on business R&D and on technology development c.f. academic research.

https://insight.jbs.cam.ac.uk/assets/Main-report-Creating-markets-for-things-that-dont-exist.pdf

3. Exploding the Myths of the UK Economy; How Soft Companies and R&D Contracts for Customers Drive the Growth of the UK Economy. David Connell and Jocelyn Probert, Centre for Business Research, Cambridge Judge Business School

105 pages. Results of extensive research on start-up and growth strategies of successful Cambridge companies in different sectors based on interviews with founders.

https://www.cbr.cam.ac.uk/fileadmin/user_upload/centre-for-business-research/downloads/special-reports/specialreport-explodingthemyths.pdf

4. Short Histories of UK STEM Based “Mittelstand” Companies

http://www.davidconnell.org/uk-mittelstand-companies.pdf

5. The Role of TICs in Rejuvenating British Industry: Submission to House of Commons Committee on Science and Technology Enquiry on Technology Innovation Centres David Connell, Professor Alan Hughes and Dr Andrea Mina. December 2010. Centre for Business Research, Judge Business School.

13 pages. Based on largely unpublished research on Intermediate Research, based on interviews with organisations in Europe, the US and Asia.

https://publications.parliament.uk/pa/cm201011/cmselect/cmsctech/writev/innovation/m19.htm

6. Innovation Myths: Game Changing Policies to Improve the Commercialisation of Research; Submission to House of Commons Select Committee Inquiry on Bridging the “valley of death”: improving the commercialisation of research. David Connell, March 2013. Centre for Business Research, Cambridge Judge Business School

11 pages. Includes an analysis of the key challenges facing academics in commercialising their work based on five year EPSRC funded research project

https://publications.parliament.uk/pa/cm201213/cmselect/cmsctech/348/348we05.htm