This is very intelligent and well intentioned research. However, we have been living the experience of this in Europe for the past few decades. It runs into serious problems. The short version of these is 3 things:
1. Governments are very bad at choosing which technologies matter and taking a long term view. Money follows an "expert" agenda which depends on the forecasts that sound best to a section of the electorate and high end media. No-one can predicting where innovation is needed least of all governments.
2. Governments are even worse at picking which funds and/ or companies to invest in. The larger the government influence, the less likely that the capital will be well allocated.
3. the whole process becomes bureaucratic and highly illegible. Government adds a burden of compliance and direction that is unworkable. Consequently a significant portion of the money is siphoned off into an ecosystem of advisors and gatekeepers that add no value to anyone.
The amounts of money involved are tiny relative to national budgets. They provide no leverage against bigger electoral priorities - PR window dressing only.
A far better approach is for Government to use its considerable spending muscle and become a good customer for innovative companies. Leave the allocation of capital to the whole spectrum of private funds: yes VCs but also Private Equity, Public Markets, Corporate investment, angels, lenders and the rest. Just because an idea doesn't work for VC, leaves space for other forms of capital. If a genuinely strong investment opportunity exists then shape a new capital instrument to serve it.
Hi Kenny, thanks for the comment and the well reasoned feedback.
1. Absolutely, and this is a key finding of the Lerner and Gompers paper cited at the top. They use In-Q-Tel as as a positive example here, showing that when government agencies are properly focused on their core capabilities and long-term strategy it can work well. So, that means, in the context of government venture philanthropy or megafunds, targetting peristent problem categories like energy, food and water security, or defence capability.
The trap is that governments are seduced by specific technology waves, such as AI, and end up overinvesting when the market is already very hot.
2. Agreed, and the main remedy here is that across all three of these layers the government is effecctively taking an index approach. That still comes with challenges around how do you add enough friction to deter opportunists, and how do you screen out obviously bad firms or companies, but the hurdle is in theory lower.
3. Well said, and this was a key finding of the Bocconi / Ifo Institute paper. The index approach helps somewhat here, as there should be an institutional tolerance for greater failure, though it's certainly a consideration for the megafund approach. I'm not sure there's a better answer than it must simply be well designed.
You're right that, where appropriate, the government is a better customer than it is a source of investment. However if they are to spur the development of new technologies that's not always possible.
Thanks for the long and considered reply - especially highlighting specifics from the papers cited which I have not read. Well designed is key but unfortunately governments don’t have much track record in that respect. Ultimately, I live in the UK and I would love to see a whole lot less government in this space - the ideal is somewhere in between as always.
Smart in principle. I think the danger lies in how it could be gamed by bad actors and fraudsters.
For example, Capital Pilot tried something relatively similar to the Venture Philanthropy part of this by offering startups an "investability test" that quickly rated their decks and then invested £50k in return for options (admittedly this was SaaS and not deeptech primarily).
Not withstanding their LP not coughing up all the cash, I think the big flaw with their model was shysters learning their criteria and then making up startup pitches to get the funds. I could see something happening like this with the model here.
The other challenge is that most university IP doesn't get funding for a reason: it's junk. Being able to sort the wheat from the chaff is still important, because if lots of junk gets funding, it will take talented human efforts away from the good stuff (in my mind anyway - perhaps this becomes less of an issue as the funding tide lifts and people can start to hop to the best opportunities).
You already highlight one measure, where a focus on more R&D intensive categories should be more resistant to oppportunism than generic software startups.
On sorting the wheat from the chaff, this is definitely a central challenge — as we know, many important ideas were initially dismissed by the creator's peers as unworkable, so there's no easy solution.
The index-like approach is designed to absorb these failures of experimentation in funding ideas that seem predictably bad in hindsight, whether they succeed or not.
This is very intelligent and well intentioned research. However, we have been living the experience of this in Europe for the past few decades. It runs into serious problems. The short version of these is 3 things:
1. Governments are very bad at choosing which technologies matter and taking a long term view. Money follows an "expert" agenda which depends on the forecasts that sound best to a section of the electorate and high end media. No-one can predicting where innovation is needed least of all governments.
2. Governments are even worse at picking which funds and/ or companies to invest in. The larger the government influence, the less likely that the capital will be well allocated.
3. the whole process becomes bureaucratic and highly illegible. Government adds a burden of compliance and direction that is unworkable. Consequently a significant portion of the money is siphoned off into an ecosystem of advisors and gatekeepers that add no value to anyone.
The amounts of money involved are tiny relative to national budgets. They provide no leverage against bigger electoral priorities - PR window dressing only.
A far better approach is for Government to use its considerable spending muscle and become a good customer for innovative companies. Leave the allocation of capital to the whole spectrum of private funds: yes VCs but also Private Equity, Public Markets, Corporate investment, angels, lenders and the rest. Just because an idea doesn't work for VC, leaves space for other forms of capital. If a genuinely strong investment opportunity exists then shape a new capital instrument to serve it.
Hi Kenny, thanks for the comment and the well reasoned feedback.
1. Absolutely, and this is a key finding of the Lerner and Gompers paper cited at the top. They use In-Q-Tel as as a positive example here, showing that when government agencies are properly focused on their core capabilities and long-term strategy it can work well. So, that means, in the context of government venture philanthropy or megafunds, targetting peristent problem categories like energy, food and water security, or defence capability.
The trap is that governments are seduced by specific technology waves, such as AI, and end up overinvesting when the market is already very hot.
2. Agreed, and the main remedy here is that across all three of these layers the government is effecctively taking an index approach. That still comes with challenges around how do you add enough friction to deter opportunists, and how do you screen out obviously bad firms or companies, but the hurdle is in theory lower.
3. Well said, and this was a key finding of the Bocconi / Ifo Institute paper. The index approach helps somewhat here, as there should be an institutional tolerance for greater failure, though it's certainly a consideration for the megafund approach. I'm not sure there's a better answer than it must simply be well designed.
You're right that, where appropriate, the government is a better customer than it is a source of investment. However if they are to spur the development of new technologies that's not always possible.
Thanks for the long and considered reply - especially highlighting specifics from the papers cited which I have not read. Well designed is key but unfortunately governments don’t have much track record in that respect. Ultimately, I live in the UK and I would love to see a whole lot less government in this space - the ideal is somewhere in between as always.
Smart in principle. I think the danger lies in how it could be gamed by bad actors and fraudsters.
For example, Capital Pilot tried something relatively similar to the Venture Philanthropy part of this by offering startups an "investability test" that quickly rated their decks and then invested £50k in return for options (admittedly this was SaaS and not deeptech primarily).
Not withstanding their LP not coughing up all the cash, I think the big flaw with their model was shysters learning their criteria and then making up startup pitches to get the funds. I could see something happening like this with the model here.
The other challenge is that most university IP doesn't get funding for a reason: it's junk. Being able to sort the wheat from the chaff is still important, because if lots of junk gets funding, it will take talented human efforts away from the good stuff (in my mind anyway - perhaps this becomes less of an issue as the funding tide lifts and people can start to hop to the best opportunities).
Agree with the concerns, David.
You already highlight one measure, where a focus on more R&D intensive categories should be more resistant to oppportunism than generic software startups.
On sorting the wheat from the chaff, this is definitely a central challenge — as we know, many important ideas were initially dismissed by the creator's peers as unworkable, so there's no easy solution.
The index-like approach is designed to absorb these failures of experimentation in funding ideas that seem predictably bad in hindsight, whether they succeed or not.