Only 10% of State and National Governments Distinguish Startups from New Businesses

A practical startup economic policy framework for governments that want venture-scale companies

Governments around the world love the word entrepreneurship. They put it in press releases, name departments after it, fund “innovation hubs,” and cut ribbons in front of coworking spaces that end up being networking clubs for service providers seeking customers. What almost none of them do (at least not formally, structurally, or coherently) is distinguish startups from new businesses. Fewer than roughly 10–15 countries worldwide do this in law, policy, or administrative design. That’s under ten percent of governments operating as if the two are economically different things. They are! Radically.

This isn’t a semantic quibble (though I know some of you will argue with me about it). It’s a policy failure with measurable consequences.

The Organisation for Economic Co-operation and Development has been explicit for more than a decade that high-growth, innovation-driven firms behave differently from small and medium enterprises and require different policy instruments, capital structures, and timelines. The World Bank draws a clear line between subsistence or lifestyle entrepreneurship and innovation-led firm formation tied to productivity growth.  Yet most governments still govern as if opening a dry cleaner or building a software development company are the equivalent of a new AI driven AdTech venture or being the first to market with gluten-free pie crust finding its way to grocery stores.

They aren’t even cousins, they’re different species.

What most states and countries do have are entrepreneurship programs, usually actually designed around job creation, self-employment, and local services. Those programs blend startups into the same category as sole proprietorships and franchises because it’s administratively easy and politically safe. The result is that startup founders are taught how to write business plans, manage cash flow, and apply for grants that were never designed for companies expected to fail fast, scale aggressively, or exit through acquisition.

I hosted today, a wonderful discussion of this with Gil Gonzales who has been tackling this in Arizona, California, and Ohio.  We’ll have the debrief of that event available soon; until then, since this topic caught your attention, register here to join me on February 3rd when I’ll be joined by Alberto Martellini and Richard Mort as we share with you how a cross-border startup program in your city opens the doors to collaboration and capital in ways that soft-landing programs fall short.

Most countries also have innovation programs, which sound promising until you look closely at the reality that innovation policy typically focuses on research outputs, patents, or technology demonstrations. Innovation is part of startups, but innovation alone does not create companies. A patent without a market is a résumé line, not an enterprise. Mariana Mazzucato, Professor in the Economics of Innovation and Public Value UCL (University College London), has written extensively about the limits of innovation policy that stops short of firm creation and market formation – such as in her work with Henry Lishi Li, Research Fellow in Health Innovation and Policy Engagement, UCL Institute of Innovation and Public Purpose, “The Entrepreneurial State and public options: Socialising risks and rewards.”

Then there’s commercialization and tech transfer, especially in university settings. Governments invest heavily here and then act surprised when the results disappoint. Technology transfer offices optimize for licensing revenue and risk avoidance, not for building venture-backed firms. According to AUTM’s U.S. Licensing Activity Survey, universities disclose tens of thousands of inventions annually, but only about 3–5% result in the formation of a startup, and AUTM does not track whether those startups survive, scale, or attract venture capital (meaning the share that become viable, venture-scale companies is necessarily much smaller (if private sector patterns are consistent that 90% of those will fail).  Commercialization is a transaction. Startups are an organizational, behavioral, and market-driven process. Confusing the two guarantees mediocrity in both.

Governments also invest billions in R&D, often pointing to this as proof of startup support. But R&D is just one input. Startups are not R&D projects; or rather, they actually are, but not in the sense meant herein, they are experiments in market creation under extreme uncertainty. Paul Graham has been explicit that startups are “companies designed to grow fast,” not labs with payroll.

Finally, there’s a question to appreciate in finance. Almost every government regulates venture capital. Some allocate capital into funds. Very few treat venture capital as a distinct economic function; the mechanism by which high-risk, non-bankable firms are created and disciplined by markets. Public capital is often deployed as grants because grants are politically palatable. Venture capital is not a grant system; it is a governance system for uncertainty. The National Venture Capital Association has been clear that venture-backed firms account for a disproportionate share of innovation, IPOs, and productivity growth in the U.S.

And did you know? Only around 50–75 universities, almost all elite or research institutions, even teach Venture Capital. Community Colleges round to zero. This means the investor class funding startups likely isn’t even exposed to how startups work unless they’ve been in one; if we’re lucky, drawing from finance degrees which are designed around traditional economics and business models.

Most governments regulate that system without understanding it.

The consequence of all this blending is predictable. Startup founders find themselves in programs optimized for stability instead of the methodology, systems, and culture of what makes startups, distinctly, work -> Universities measure patents instead of companies; public capital substitutes for private discipline; policymakers declare success because activity happened, not because firms scaled… And then everyone wonders why ecosystems stagnate.

Countries that do distinguish startups from new businesses (Israel, Estonia, France, Finland, Singapore, South Korea, the UK, Germany, Chile, Canada, Australia, Japan) govern differently. France’s La French Tech explicitly separated startups from SMEs across labor law, taxation, and capital access.  Israel’s entire innovation policy framework was designed more with venture formation in mind, not small-business support. The outcomes are obvious.

If governments actually want startups (not press releases about them) some policy shifts are unavoidable

A few thoughts…

First, economics and marketing must be core curriculum everywhere, not electives and not optional. Marketing is not promotion; it is value is discovered and created. Peter Drucker was unambiguous: “The aim of marketing is to know and understand the customer so well the product or service fits him and sells itself.” Startups fail because they misunderstand incentives, demand, and behavior, not because they can’t code. Teaching STEM without economics is how you get elegant solutions to nonexistent problems.

Second, startup-specific committees should exist in every legislative body and major department. Not “small business.” Not “innovation.” Startups. Policy design requires fluency. You wouldn’t ask an agriculture committee to regulate aerospace; treating startups as generic businesses is the same category error.

Third, governments should fund startup infrastructure (physical and platform). Free or heavily subsidized spaces for coworking, events, and founder collisions outperform almost every curriculum-driven intervention. Density matters. MIT’s research on innovation districts shows proximity and interaction (not subsidies) drive entrepreneurial output.  Spaces should be neutral, open, and boring in governance.  Along with that, the platform on which to run startup development programs, the CRM of investors, and the community networking software – who is going to pay for that when startups can’t (and founders shouldn’t)?  Or rather, why make everyone reinvent the wheel when the wheels are in motion and available??

Fourth, grant programs, such as the United States’ SBIR and STTR, must be explicitly startup-oriented, not research-oriented. Not in NAME, in practice, which is to say, clear carve out of what is available, how, and why, for startups rather than everything else bundled together.  In the U.S., these programs are still evaluated as science projects despite repeated evidence that commercialization outcomes improve when startups (not institutions) are the unit of analysis.  If a grant isn’t structured to lead toward a venture-scale company, it shouldn’t be counted as startup policy.

Fifth, healthcare (especially mental health coverage) must be accessible to startup founders independent of employment status (hint, they’re not employees when a temporary venture is neither funded nor making money). The data on founder stress, depression, and anxiety is no longer anecdotal. Michael Freeman’s widely cited study found founders are significantly more likely to experience mental health challenges than the general population. Treating that as a personal failing instead of a systemic risk is negligent.

Three more changes come immediately to mind…

Governments need procurement pathways designed for startups, with faster timelines and smaller contracts. The United States Department of War has shown this is possible through programs like AFWERX and Defense Innovation Unit (DIU), which explicitly target startups rather than incumbents.

They need immigration policies that treat founders as economic infrastructure, not labor market anomalies. Canada’s Startup Visa and Estonia’s Startup Visa exist because those governments understand that founders import future firms, not just talent.  What caused Silicon Valley to boom?  Migration into Northern California just as would happen to Austin, TX a couple decades later.  Immigration brings culture, different opinions, other experiences, and networks, into the ecosystem.

And they need data systems that track startups separately from small businesses, measuring formation, failure, follow-on capital, and exits. You cannot manage what you refuse to define. Kauffman Foundation research has repeatedly shown that high-growth firms drive net job creation, yet most state dashboards still lump them into SMB metrics.

None of this is radical. What’s radical is continuing to pretend that startups are just enthusiastic small businesses with hoodies.

If you’re a policymaker, economic developer, or university administrator reading this and feeling defensive, that reaction is the point. Ask yourself a simple question: where, exactly, in your statutes, budgets, or org charts does “startup” appear as a distinct category?  Not a buzzword meaning new businesses but rather a distinction of your plans and policies different from new business creation.  If the answer is nowhere, you’re not supporting startups, you’re crowding them out with good intentions.

And if you’re a founder operating inside that system, the better question is which of these absences is slowing you down right now (and why you’re expected to absorb that cost while governments celebrate “entrepreneurship,” and the jobs created that are a result of your hand).