On 7 November New Zealanders will vote to elect a new government. Just about every party contesting this election agrees we need to do something different to lift our productivity and change the trajectory of our economic performance. And when they talk about how, startups and innovation are almost always part of the proposed solution.
This is nothing new. Aspiring politicians have been saying similar things my entire life. Meanwhile, we’ve kept falling further behind.1
What could the next government actually do differently?
Policy ideas typically start with the question: “what can the government do for startups?” I think the more useful question is the reverse: “what can startups do for the government, and so for everyone?”
It’s easy to think of new and different ways to give money away. It’s much harder to identify the return on investment. Any proposal should start with the people it is aiming to serve and work backwards to the policy that will measurably achieve that result.
There are four questions to be answered ahead of funding:
Who does this help?
What constraints do they actually have?
How will this reduce or remove those constraints?
How will we show it’s working?
Evidence of impact means early indicators, milestones and a control group, so any result can be honestly attributed. If there’s no way to show it’s working, it isn’t an experiment; it’s just wishful thinking.
Here are three ideas that hold startup policy to that standard, drawn from my book How To Be Wrong ...
1. Choose one metric
We start with a mish-mash of competing objectives.
We say we want more early-stage companies, more high-growth companies, to attract capital from overseas and at the same time to retain local ownership (we have two different government agencies working on these two competing objectives today!), more highly paid jobs, more tax, more exports, more imported capital, more recycled capital, to attract “talent”, and to grow wealth. We have programmes in place to incentivise all of these things, and mostly don’t know if they are working or not.
Until the “why” is clearer, prioritising and assessing the “what” is impossible.
For me this is an easy question. Policy should be judged by what it returns to those paying for it.
Be honest about the collective value of successful startups.
Startups are rarely profitable, so they don’t pay company tax. And there is no capital gains tax when shareholders make large profits when companies are sold. The capital that successful local founders and employees recycle into the next generation of ventures is vital to growing the ecosystem, but doesn’t benefit the government directly. Decades of evidence show that here a rising tide doesn’t lift all boats. Wealth doesn’t automatically trickle down.
The only direct contribution to government from startups is PAYE on local salaries and GST on local operating expenses.
Given that, the obvious goal should be to increase the number of highly paid jobs these companies create in New Zealand.2
We should track and report on the number of new highly paid jobs, and the PAYE they generate, relative to the money we spend incentivising startups in their early stages, so all of us can be confident there is a return on our investment.
Be honest about the actual constraint. It’s people, not capital.
Venture returns follow a power law: most early-stage investments lose money, the rare winner dwarfs everything else, and only the small number of companies that reach escape velocity ever pay back the money that is invested in getting them off the ground. Targeting the long tail is not a path to success. Treating startups as a lottery and buying as many tickets as possible doesn’t increase the odds of a win; it just raises the amount spent.
The thing that’s genuinely scarce is fundable companies. And when we ask them specifically they nearly all have the same fundamental constraint: people.
In 2011 Sir Paul Callaghan famously said: “I believe that just a hundred inspired entrepreneurs could turn this country around. It only takes one genius entrepreneur to make a company like this. So a hundred individuals could earn us $40 billion a year in exports.”
He was wrong. We now know from experience that every one of those founders needs to hire hundreds, often thousands, of qualified specialists to help them grow their companies. We don’t just need 100 inspired founders, we need 10,000 skilled contributors, then 100,000 more.
That makes immigration one of the biggest short-term levers we have. Our current settings import the wrong thing. The Active Investor Plus visa imports capital, not operators. Recently this scheme has been successful in attracting billions of dollars, but most of that money has been invested in bonds and managed funds, rather than directly into high-growth companies. Attracting people at the end of a successful career and encouraging them to spend a short period of time each year in New Zealand is like a second-division French rugby club signing a soon-to-retire All Black for one final season and expecting it to transform their team’s performance.
We should orient the system toward people still hungry to build. There should be a fast, low-friction pathway for experienced operators who want to come to work. And, at the same time, we should be actively recruiting experienced New Zealanders already here out of low-productivity corporate and bureaucratic roles into teams where they could really make an impact. We should be building genuinely diverse teams to widen the available pool.
At the moment our tax settings treat the people who build these companies worse than the people who fund them. While investors pay 0% capital gains tax, employee equity is taxed as ordinary income, up to 39%. While the recent changes make it easier for this tax to be deferred (ensuring people are not left paying tax before they have received the income) they don’t change the rates. Ultimately, the only way to close this gap is to charge investors more or employees less.
Of course, the longer-term investments matter too and should run in parallel. We need to invest in training far more people in areas where there are skill gaps. At the moment the number of graduates in science, engineering and IT degrees is insufficient to even cover growth in demand, let alone the backlog of open roles. That all but guarantees our productivity will continue to drift.
Who does this help? New Zealanders in highly paid jobs, and the taxpayers who fund startup support and get their return through the PAYE and GST those jobs generate.
What constraints do they actually have? Ambitious companies can’t hire enough experienced people to fuel their growth.
How will this reduce or remove those constraints? Immigration settings that import operators rather than retirees, tax settings that treat the people building companies at least as well as the people funding them, and training to close the gap over the longer term.
How will we show it’s working? We already collect the data. Count the highly paid jobs created by the companies we support, and the PAYE they generate, against what that support costs.
2. Each to their own
If we’re determined to fund startups, we should be clear about the roles that government plays.
What are the things that only government can do? Here are three, for starters:
It can deploy funding at scale;
It can collect the data needed to target that funding; and
It sets the rules that everybody else operates within.
What it has proven to be bad at, again and again, is choosing who should get that funding.
Investors and founders are the opposite. Choosing is the whole job. Those of us who do this live with the consequences when we get it wrong.
So, the question we should ask when setting policy is how to let government do the funding without doing the choosing. The answer, I think, is to link money to results, not applications.
Today we have a complicated system of multiple departments and agencies overseeing a fragmented array of government grants and funds. For founders this is time-consuming and frustrating. There should be a single point of contact for companies who want support, applied consistently, with as little subjective bureaucracy as possible - ideally none.
Deciding company-by-company who deserves a grant puts officials in an impossible position.
If they try to pick winners they expose their own biases. If they try to fund everyone they waste money on failures. Either way it produces a large, expensive apparatus. It distracts the founders it’s meant to help with hoop-jumping. It distorts behaviour, as companies learn to describe what they were going to do anyway in the language of whatever grants will pay for: building the next version of the product becomes an “R&D project”, hiring a sales lead becomes “market development”. And it spawns a cottage industry of consultants, often ex-agency staff, who know the right keywords for the right funding application forms.
The work doesn’t change, only the paperwork does. That’s the opposite of impact.
Pay for the outcome, not the inputs.
When we provide funding to early-stage and high-growth companies we should pay for the thing we actually want, which is more highly paid jobs based in New Zealand. Not the proxies we hope lead there, such as R&D spend, and definitely not vanity metrics like the number of startups created or capital raised.
Productivity and exports matter, but as signals rather than separate goals. A company can only keep paying high salaries if each person generates a lot of revenue, so revenue per full-time employee is the useful definition of productivity. And in a market our size, the companies that can keep adding those jobs are almost always the ones selling overseas. So export revenue per dollar of payroll is the countermeasure. It checks that those jobs are real and paying their way, rather than being a target in its own right.
The much harder question is how to do that objectively, without anyone having to judge each application.
The government already has all of the data needed to allocate funding objectively.
We should use data the government already collects. Every employer files PAYE returns which detail who is paid what, and every GST-registered business files GST returns which detail local spending and overseas revenue. From that existing system we could objectively identify exporting companies, and use the attributes we want to encourage to decide who qualifies: fast-growing rather than mature, higher revenue-per-employee, smaller rather than larger, majority local ownership, or any combination of these things.
For example, qualifying companies could automatically receive a top-up on two things they already report: a rebate on the PAYE they pay for salaries above a published threshold, and a multiplier on their existing monthly GST refund, which reflects their local spending. These would be small and regular payments, not lumpy annual grants, paid directly into a bank account the government already has on file. The criteria could be published in advance so the policy can be debated on its merits; every company can work out its own entitlement; the payment scales with success, so money flows to companies that are producing the desired results rather than to early-stage failures; and there is nothing for an official to subjectively approve.
Yes, some of those jobs would have been created anyway. That’s true of every grant scheme. The difference is that this one is measured, and the companies just below the threshold show us what it actually changed.
This is what public funding for startups would look like if we acknowledged the power law, rather than fighting it.
Building a new bureaucracy isn’t the answer.
Shutting Callaghan Innovation was the correct decision. We don’t need to debate this. We have eleven years of evidence, and the numbers are eye-watering. Between 2013 and 2025 it distributed $1.7 billion but spent another $1 billion overseeing that funding. That is a terrible ratio, which should have been questioned more at the time.3
However, scattering the functions of that single agency across multiple agencies has made the problem worse. Not only is it now more or less impossible to measure, it’s made it even more confusing and time-consuming for founders trying to access support.
Ironically, given the things to which we’ve subsequently attached his name, back in 2011 Sir Paul Callaghan explicitly warned against picking favourites. He said that these businesses would be unpredictable. He anticipated that our competitive advantage would be in obscure niches, and that the companies that were eventually successful would surprise us. He was right about all of those things.
We should acknowledge that trying to pick winners is a failed policy. Objective, automatic support is cheaper to run, harder to game, and easier to defend.
For all of the same reasons we should also gracefully wind down NZ Growth Capital Partners (NZGCP). This was established way back in 2002 (as NZ Venture Investment Fund) in an attempt to replicate the successful Israeli “Yozma” model and bootstrap a venture capital sector. That failed miserably (as I described in the book, by 2018 they reported -3.29% returns on capital invested). Since then we’ve continued to double down on that strategy, hoping for different outcomes.
While more recent results have seemingly been better, in the meantime, a healthy and growing venture capital sector has formed, not on the back of government subsidies but as the exhaust of successful ventures - for example, some of the proceeds from the Trade Me sale were recycled as seed funding for ventures like Xero and Vend, and to new venture funds like Movac. This is exactly how we should expect it to work.
Thanks to the growth of money managed within the KiwiSaver scheme, we can anticipate that local venture funds who demonstrate good judgement will continue to attract even more capital. And likewise, we don’t need to fear for the NZGCP team - those who add value to the companies they invest in will quickly find roles in these growing private sector funds.
It’s time for the government to get out of the way and stop distorting this market.
Who does this help? Companies already creating highly paid jobs here, and through them the taxpayers funding the support.
What constraints do they actually have? A support system that is expensive to access and rewards writing applications over getting results.
How will this reduce or remove those constraints? Eligibility and payments are calculated from returns companies already file, so there is no application to write and nothing for an official to administer or approve.
How will we show it’s working? Published thresholds create a natural control group: the companies just below the line. If qualifying companies add highly paid jobs faster than those just short of it, the policy is working.
3. Reframe the debate
The debate on how government can support “the startup ecosystem” is stuck on three lazy assumptions. Unlike the first two ideas, these aren’t policies to test against the four questions. They’re the assumptions that stop us asking those questions in the first place.
“Tech” is not a sector.
The familiar claim that technology is our “third-largest export sector” rests on unverifiable, self-supplied data (ref: The TIN Report, prepared for many years by a private company, with generous public sector funding). However, “technology” isn’t a sector in any official Stats NZ classification - it’s an arbitrary bundle of unrelated industries.
I suggest we go much further than that ...
In 2026 every company is a technology company. Calling a startup a “tech company” is about as useful as saying the local café is in the electricity sector because that’s what powers their coffee machine.
We should stop trying to draw a line around “the sector” and instead acknowledge the goal is to back productive, exporting firms in every industry. They should all be using technology to fuel their growth.
We already have arguably the most capital-friendly tax system in the world.
We have effectively 0% capital gains tax on investments in startups. This is remarkable. In every country that we frequently compare ourselves to, investors in early-stage companies expect to pay tax on realised gains.
We could do much more to promote this advantage. And we should politely push back on any investor who feels like our system could be even more generous when they argue they need further incentives to invest.
We are not a small country a long way from anywhere.
New Zealand is large but lightly populated. As a result our genuine competitive advantage is connectedness. In a population this small, six degrees of separation collapses to two or three, which means low corruption, high trust and unusual ease of doing business. It’s harder to hide, or to be left behind out of sight, and we’re less reliant on the social codes that matter elsewhere, like where you went to school or what job you have. We can just ask someone who knows what a person is really like.
This should be celebrated more. It’s one of the key reasons why talented people will choose to live and work here.
We are also not as “remote” as we often claim. It doesn’t take weeks on a boat or hours on a plane to get data to the other side of the planet. What’s more, there is now a large and very successful diaspora of New Zealanders all around the world, who are generally very supportive of New Zealand founders and companies. We should be doing much more to tap into this group and ensure they remain connected to home - physically and emotionally.
Finally, we should stop thinking of Aotearoa as just a small country and feeling like our size is a weakness. We are a big small country. Perhaps the biggest small country! Acknowledging that would at least force us to find a better aspiration than “punching above our weight”.
From the biggest, the most is expected, right?
tl;dr
We should choose one metric to focus on: highly paid jobs based in New Zealand. It’s currently the only direct return government gets from startups, and it’s also where those companies hit their biggest constraint. Let government do the funding and startups do the choosing, by linking money to results rather than applications. And reframe the debate, so we stop wasting precious time and money on constraints that are imagined.
Now, who can we vote for that would actually do this...?
Triage: first aid for all your inboxes
Triage is first aid for everybody who struggles to stay on top of their inbox. It doesn’t try to replace traditional email software, but lets you use your downtime to quickly remove the noise and stress.
It’s an app that runs on your iPhone. New messages appear as a familiar “stack of cards”. With a flick or a tap, you can quickly archive messages, keep important ones in your inbox for later, forward to somebody else, or compose a short reply.
Triage supports Gmail, iCloud Mail, Fastmail and most other email services.
This is a tool I built first for myself, and it’s still the first app I open every morning. I’m delighted that so many others find it useful too. In the latest version, see messages from all of your accounts in one stack, and access quick actions with a long press. I hope to write more about this soon. But in the meantime, check it out and let me know what you think…
Triage v2.3 is available now on the App Store.
Subscribe for free to receive new posts
and get new posts delivered directly to your inbox.
References
How To Be Wrong: A crash course in startup success, Published February 2025 by Electric Fence.
Picking Winners, first published by The Spinoff, 2017.
M3: The Metrics Maturity Model, first published on rowansimpson.com
Why Govt shouldn’t be picking start-up winners, TVNZ Q+A, 2025
Callaghan, Rebooted, AUT Dean’s Lecture, 2025
Callaghan Innovation Annual Reports, 2013-2026
Other references for the figures in this post: https://howtobewrongbook.com/references
Compared to the top half of the OECD, New Zealand’s labour productivity gap widened from 34% in 1996 to around 40%. Source: NZ Productivity Commission (2023), via BERL, Relooking at our productivity gap, January 2025.
Arguably a capital gains tax is an even more obvious way that startups can contribute to government. But requires change that politicians have found impossible, even when they had a majority in parliament. The advantage of the goal I’ve described is that it’s achievable with the current tax system already in place. But, important to note we remain a massive outlier compared to just about every other country on this point, and this shouldn’t be an either/or, it should be a yes/and.
Some of us tried, and were often ostracised as a result. Too often the excuse of “commercial sensitivity” was used to avoid scrutiny. That’s a topic for another whole post at some point.
Thanks Rowan. 'Grow innovation' is used by many (or just me) a bit like 'the economy' . We know it's important, and have fragments of insight or data or observation but don't have any real depth of understanding about it. This framework is really useful - your last question tho, I'm not sure there is anyone who is talking in this way??