Do winners pick government? How scale-up experience shapes entrepreneurs’ assessments of innovation policy mixes

https://doi.org/10.1093/scipol/scad030

Steven Denney, Travis Southin, David A Wolfe

Bottom Line Up Front

Scaleups prefer grants to tax incentives and want government to ‘pick winners’ instead of spreading out the love. They’re positive and negative about everything.

Summary

Analyze interviews with entrepreneurs from Canadian technology firms. Is there a disconnect between objectives and instruments employed by the government?

Entrepreneurs prefer demand-side, direct, and targeted innovation instruments.

Findings provide a more nuanced understanding of the innovation policy landscape and preferences of tech scale-ups.

I’m debating staying in Toronto or going to San Francisco to create my startup.

Introduction

Winner takes most model means tech giants suck up both talent and smaller firms.

Government must support companies with the potential to scale. The comparative advantage comes from the innovative capacity of a nation.

Canada has:

  • strong startup ecosystem
  • commitment by the feds to support high-growth firms
  • performance worse than competitor countries

Canada is below average in:

  • R&D expenditure (BERD)
  • firm-level productivity
  • investments in technology
  • tech exports
  • patenting
  • startups reaching scale-up or high-growth status

Three decades of “Low innovation equilibrium”

Innovation policy mixes

3 kinds:

  1. Neutral vs targeted
  2. supply vs demand-side
  3. direct vs indirect

They give a big table, but basically, SR&ED credits are supply, indirect, and neutral. NRC labs are supply, direct, and targeted. Innovative Solutions Canada is demand-side, direct, and targeted.

Canada and France are unusual in using so much tax-based support instead of direct grants. SR&ED is as much as 90% of R&D. The fact that BERD is so low shows this isn’t working. Established firms benefit more from so-called “neutral” policies. Mostly large multi-nationals that aren’t headquartered in Canada. The few direct grants we have are for manufacturing, especially automotive and aerospace.

Nortel and RIM failed to leave no anchor firms.

ISED wanted to double Canada’s high-growth firms between 2019 and 2025, not only is it not on track, but there are fewer in 2023 than in 2019.

Scaleups are too big to be helped by incubators, and too small to take advantage of “neutral” policies which are mostly consumed by gigantic multinationals. They have a disproportionate economic impact, especially:

  • net job creation
  • productivity gains
  • innovation
  • R&D
  • exports

Startups also have high growth, but high failure rates. Scaleups typically grow but do not fail.

This feels like a bit of a circular definition only possible in hindsight. In Canada there are ‘chosen winners’ like Element AI that raised $100m+ USD seed rounds, were given millions in grants from the government with every level of government personally putting Trudeau in their offices and they still, by any reasonable measure, hired hundreds of employees, and fired over 300 employees when it got acquihired out of existence. Is this a failure? Or does a 500+ employee firm operating in multiple countries not count as a startup?

Criticisms are surprisingly concentrated in the same areas as preferences. This is explained away basically by saying a desire for greater procurement is coded as positive but frustration with the current system is coded as bad. This feels pretty weak.

Excessive reliance on tax credits is complained about. Then entrepreneurs say that picking winners is good, the government should set up some court to decide who will win. This is followed by saying that they should get more direct grants instead of tax breaks, and finally bizarre fantasy land nonsense about having CPP dump all its money into big PE buyouts of (presumably their) companies.

They point to Bombardier, Nortel, and Blackberry as examples of winners. They don’t want to grow successful companies, they want to be government darlings.