AI and Entrepreneurship: What is Really Happening?

Money flowing into startups globally fell by a third in 2022 as valuations declined; however, there is still $300B of VC dry powder in the US alone.

“…The tech-heavy Nasdaq index fell by a third in 2022, making it one of the worst years on record and drawing comparisons with the dotcom bust of 2000-01. According to the Silicon Valley Bank, a tech-focused lender, between the fourth quarters of 2021 and 2022, the average value of recently listed tech stocks in America dropped by 63%. And the plunging public valuations dragged down private ones. The value of older, larger private firms (“late-stage” in the lingo) fell by 56% after funds marked down their assets or the firms raised new capital at lower valuations. What new VC funding there is increasingly flows into mega-funds. Data from PitchBook, a research firm, show that in America in 2022 funds worth more than $1bn accounted for 57% of all capital, up from 20% in 2018….”

Economist Staff, “How the titans of tech investing are staying warm over the VC winter,” The Economist, February 26, 2023, https://www.economist.com/business/2023/02/26/how-the-titans-of-tech-investing-are-staying-warm-over-the-vc-winter

How the titans of tech investing are staying warm over the VC winter

Venture capitalists are not known for their humility. But the world’s biggest investors in innovation have been striking a more humble tone of late. In a recent letter to investors Tiger Global, a hedge fund turned venture-capital (vc) investor, reportedly admitted that it had “underestimated” inflation and “overestimated” the boost the pandemic would give to the tech startups in its portfolio. In November Sequoia, a Silicon Valley vc blue blood, apologised to investors in its funds after the spectacular blow-up of ftx, a now defunct crypto-trading platform that it had backed. Speaking in January, Jeffrey Pichet Jaensubhakij, the chief investment officer of gic, one of Singapore’s sovereign-wealth funds, said that he was “thinking much more soberly” about startup investing.

The vc giants’ newfound contrition comes on the back of a gigantic tech crash. The tech-heavy nasdaq index fell by a third in 2022, making it one of the worst years on record and drawing comparisons with the dotcom bust of 2000-01. According to the Silicon Valley Bank, a tech-focused lender, between the fourth quarters of 2021 and 2022, the average value of recently listed tech stocks in America dropped by 63%. And the plunging public valuations dragged down private ones (see chart 1). The value of older, larger private firms (“late-stage” in the lingo) fell by 56% after funds marked down their assets or the firms raised new capital at lower valuations.

This has, predictably, had a chilling effect on the business of investing in startups. Soaring inflation and rising interest rates made companies whose promised profits lie primarily in the distant future look less attractive today. Scandals like ftx did not help. After a decade-long bull run, the amount of money flowing into startups globally declined by a third in 2022, calculates cb Insights, a data provider (see chart 2). In the final three months of 2022 it fell to $66bn, two-thirds lower than a year earlier; the number of mega-rounds, in which startups raise more than $100m, fell by 71%. Unicorns, the supposedly uncommon private firms valued at more than $1bn, became rarer again: the number of new ones contracted by 86%.

This turmoil is forcing the biggest venture investors—call them the vc whales—to shift their strategies. For Silicon Valley, it signals a reversion to a forgotten style of venture capitalism, with fewer deep-pocketed tourists splashing the cash and more bets on young companies by Silicon Valley stalwarts.

Misadventure capital

To understand the scale of vc’s reversal of fortune, consider its earlier bonanza. Between 2012 and 2021 annual global investments grew roughly ten-fold, to $638bn. Conventional vc firms faced competition from a new breed of investor from beyond Silicon Valley. These included hedge funds, the venture arms of multinational companies, from Shell to Samsung, and the world’s sovereign-wealth funds, some of which began investing in startups directly. Dealmaking turned frenetic. In 2021 Tiger Global inked almost one new deal a day. Across vc-dom activity “was a bit unhinged”, says Roelof Botha, boss of Sequoia Capital, “but rational”, given that low interest rates meant that money was virtually free. And “if you weren’t doing it, your competitor was.”

What passed for rationality in the boom times now looks somewhat insane. The downturn has spooked the vc funds’ main sources of capital—their limited partners (lps). This group, which includes everyone from family offices and university endowments to industrial firms and pension funds, is growing more nervous. And stingier: lower returns from their current investments leave lps with less capital to redeploy, and collapsing stockmarkets have left many of them overallocated to private firms, whose valuations take longer to adjust and whose share of some lps’ portfolios thus suddenly exceeds their quotas. Preqin, a data firm, finds that in the last quarter of 2022 new money flowing into vc funds fell to $21bn, its lowest level since 2015.

What new vc funding there is increasingly flows into mega-funds. Data from PitchBook, a research firm, show that in America in 2022 funds worth more than $1bn accounted for 57% of all capital, up from 20% in 2018. How the vc whales behind these outsize pools of capital adapt to the vc winter will determine the shape of the industry in the years to come.

The venture cetaceans can be divided into three big subspecies, each typified by big-name investors. The startups they finance range from the newly founded, in need of “seed” funding, to the somewhat older, later stage firms that are looking to rapidly grow. First there is the conventional Silicon Valley royalty, such as Sequoia and Andreessen Horowitz. The second group comprises the private tourists, such as Tiger and its New York hedge-fund rival, Coatue, as well as SoftBank, a gung-ho Japanese investment house. Then there are the sovereign-wealth funds, such as Singapore’s gic and Temasek, Saudi Arabia’s Public Investment Fund (pif) and Mubadala of the United Arab Emirates. As well as investing directly, these entities are lps in other vc funds; pif, for example, is a large backer of SoftBank’s Vision Fund.

Together these nine institutions ploughed more than $200bn into startups in 2021 alone, or roughly a third of the global total (not counting the state funds’ indirect investments as lps). All nine have been badly damaged by last year’s crash. Sequoia’s crossover fund, which invested in both public and private firms, reportedly lost two-fifths of its value in 2022. Temasek’s listed holdings on American exchanges shrank by about the same. SoftBank’s mammoth Vision Funds, which together raised around $150bn, lost more than $60bn, wiping out their previous gains. In a sign that things were terrible, its typically garrulous boss, Son Masayoshi, sat out its latest earnings call on February 7th. Tiger reportedly lost over half of the value of its flagship hedge fund and marked down its private investments by about a quarter, torching $42bn in value and leading one vc grandee to speculate that the hedge fund might turn itself into a family office.

All three groups have reined in their investments. But each has responded to the downturn in distinct ways. That is in part because it has affected them to different degrees.

The private outsiders have been hardest hit. The combined number of startup investments by the three firms in our sample fell by 76% between the second half of 2021 and the same period in 2022. Tiger has lowered the target for its latest fund from $6bn to $5bn; its previous one raised $13bn. In October Phillipe Laffont, Coatue’s boss, said that the hedge fund was holding 70-80% of its assets in cash. The firm has also raised $2bn for its “tactical solutions fund”, designed to give mature startups access to debt and other resources, as an alternative to raising equity at diminished valuations during a market downturn. SoftBank has all but stopped investing in new startups. Instead, in the second half of 2022 most of its capital went to well-performing portfolio firms, says Lydia Jett, a partner at the Vision Fund.

The other two groups are also retrenching, but not as drastically. According to data from PitchBook, in the second half of 2022 the number of deals struck by Sequoia and Andreessen Horowitz fell by a combined 47%. Direct investments by the four sovereign funds in our sample slowed by a more modest 31% in the same period, no doubt thanks to their governments’ deep pockets and longer time horizons.

Taken together, venture capitalists’ slowing pace of investment has left them with a record amount of capital that lps had already pledged to stump up but that has yet to be put to use. Last year the amount of this “dry powder” was just shy of $300bn in America alone (see chart 3). According to data from PitchBook, our five private whales are sitting on a combined $50bn or so; the sovereign investors hold their numbers close to their chest but are likely to be of a similar order of magnitude, all told. Some of it may wait a long while to be deployed, if it ever is. But some will find grateful recipients. Who those recipients are also depends on which group of whales you look at.

Conventional vcs and the hedge funds are focusing on younger “early-stage” firms. in part because volatility in the public markets makes it harder to value more mature companies that hope to list in the near future. Mr Botha says that Sequoia has doubled the number of “seed” deals with the youngest companies in 2022, compared with 2021. In January the firm launched its fifth seed fund, worth $195m. Last April Andreessen Horowitz launched an “accelerator” programme to nurture startups. About half the startups Tiger backed in 2022 were worth $50m or less, compared with just a fifth in 2021, according to PitchBook.

Early-stage firms are unlikely to be the only recipient of vc cash. David DiPietro, head of private equity at T. Rowe Price, a fund-management group, thinks that startups selling “must-have” products, such as cyber-security services, or cost-cutting tools, such as budgeting software, should fare well. Money will also keep flowing to well-managed businesses with strong balance-sheets., expects Kelly Rodriques, chief executive of Forge, a marketplace for private securities. Firms with buzzy new technologies, such as artificial-intelligence chatbots and other forms of whizzy “generative ai”, are also likely to attract investments—especially if those technologies already work in practice and underpin a viable business model.

Another category of startups likely to gain favour comprises those involved in industries that governments deem strategic. In America, that means climate-friendly technology and advanced manufacturing, on which Uncle Sam is showering subsidies and government contracts. Some 8% of the deals all our whales made in the second half of 2022 involved firms working on technologies to combat climate change, for example, up from 2% in the same period of 2021. Last year Andreessen Horowitz launched an “American Dynamism” fund, which partly invests in firms that rely on government procurement, such as Anduril, a defence-tech startup.

Sovereign-wealth funds are likely to be looking elsewhere. Seed deals are simply too small for them: whereas the typical early-stage American company is worth about $50m, in 2021 the median value of startups backed by the sovereign funds was a whopping $650m. And to them, what counts as “must-have” startups is somewhat different, determined less by the market or other states’ strategic imperatives, and more by their own governments’ nation-building plans.

On February 16th pif said it would take a stake in vspo, a Chinese platform for video-game tournaments. This is part of a plan dreamed up by Muhammad bin Salman, the Saudi crown prince, to invest $38bn in “e-sports” by 2030 and make Saudi Arabia a gamer’s mecca. Temasek invests heavily in firms that develop technology to boost food production. In the past year it backed Upside Foods, a startup selling lab-grown meat, and InnovaFeed, a maker of insect-based protein. This is motivated by Singapore’s goal of locally producing 30% of the city-state’s nutritional needs by 2030, up from about 10% in 2020. Rohit Sipahimalani, chief investment officer of Temasek, thinks that over the next few years his focus will shift towards “breakthrough innovation rather than incremental innovation”, on the back of government support of strategic tech.

One group of firms is likely to see less investment from our whales, however: those in China. The Communist Party’s harsh two-year crackdown on consumer technology may be easing but the vc titans remain wary of what was until recently one of the world’s hottest startup scenes. An executive at a big venture fund says that in the past, foreign investors in China knew that the government would be respectful of their capital. Now, he sighs, it feels like the government “has pulled the rug out from underneath us”.

Tiger has said that there is a “high bar” for new investments in China. gic has reportedly scaled back its investments in China-focused private funds. Mr Sipahimalani of Temasek says diplomatically that he is trying to avoid investing in “areas caught in the cross-hairs of us-China tension”. Sequoia is reportedly asking external experts to screen new investments made by its Chinese arm into quantum computing and semiconductors, two such contentious areas. All told, the number of our whales’ deals with Chinese startups fell from 22% of the total in 2021 to 16% in 2022.

After the dotcom crunch vc investments needed nearly two decades to return to their previous peak. Today’s tech industry is more mature, startups’ balance-sheets are stronger and, according to the Silicon Valley Bank, their peak valuations relative to sales are lower than in 2000-01. This time the whales of vc are unlikely to need 20 years to nurse their wounds. But the experience will have lasting effects on the sort of businesses they back.

According to data from Insider Intelligence, Google and Facebook’s share of digital advertising was 48.4% in 2022, and is expected to decline to 44.9% in 2023, as Amazon, TikTok, and digital streamers gain share.

“…For the first time in nearly a decade, the two largest players in online advertising are no longer raking in the majority of U.S. digital-ad dollars, a decline that industry insiders expect to continue in years to come. Alphabet Inc.’s; Google and Facebook parent Meta Platforms Inc. accounted for a combined 48.4% of U.S. digital-ad spending in 2022, according to estimates from research firm Insider Intelligence Inc. Their combined U.S. market share hadn’t been under 50% since 2014, said Insider Intelligence, which expects that number to drop to 44.9% this year….”

Patience Haggin, “Google and Meta’s Advertising Dominance Fades as TikTok, Streamers Emerge,” Wall Street Journal, January 4, 2023, https://www.wsj.com/articles/google-and-metas-advertising-dominance-fades-as-tiktok-netflix-emerge-11672711107

Google and Meta’s Advertising Dominance Fades as TikTok, Streamers Emerge

For the first time in nearly a decade, the two largest players in online advertising are no longer raking in the majority of U.S. digital-ad dollars, a decline that industry insiders expect to continue in years to come.

Alphabet Inc.’s Google and Facebook parent Meta META Platforms Inc. accounted for a combined 48.4% of U.S. digital-ad spending in 2022, according to estimates from research firm Insider Intelligence Inc. Their combined U.S. market share hadn’t been under 50% since 2014, said Insider Intelligence, which expects that number to drop to 44.9% this year.

The ad businesses of Google and Meta are still growing, but Insider’s data suggest the pace is slower than the rest of the U.S. digital-advertising market. Insider forecasting analyst Zachary Goldner said the erosion of their combined market share was the result of brands having access to more advertising formats.

“All marketers want more options,” Mr. Goldner said.

Google and Meta each faced headwinds in 2022, as people spent less time online than in the early days of the pandemic; marketers concerned about a possible economic downturn reined in ad spending; Amazon and ByteDance Ltd.’s TikTok continued their emergence as a force in digital advertising; and more streaming services started to embrace advertising.

Meta and other social-media companies including Snap Inc. also suffered from Apple Inc.’s 2021 decision to require apps on its devices to ask users if they wanted to be tracked. The majority of iPhone users opted not to be, hitting the heart of Meta’s business: its ability to target ads at users with precision and prove to marketers that the ads generate sales.

Google wasn’t as affected by Apple’s move, because its flagship search-ad business relies on customer intent—users’ search terms immediately reveal what they are interested in—rather than data collected from app and web tracking. Its U.S. digital-advertising market share slightly rose to 28.8% in 2022, Insider Intelligence said, but is expected to fall to 26.5% this year.

Google didn’t respond to a request for comment. On the company’s most recent earnings call in October, executives talked about how search-based advertising tends to do well during challenging economic times.

Meta had no comment. In its most recent earnings call, Meta said the tracking restriction continued to affect its ad business but noted the effect had diminished.

Apple’s tracking restriction had an impact for emerging e-commerce companies, which are important to Meta’s ad business. Supergut Chief Executive Marc Washington said the maker of gut-health products used to spend about 80% of its ad budget on Meta’s Facebook and Instagram platforms, with the rest going to Google. In early 2022, he noticed that the cost of bringing in new customers through advertising on Meta’s platforms was twice as high as it was before Apple’s privacy changes. Supergut shifted about half of what it spent on Meta to TikTok, a short-form video platform popular with younger audiences.

Mr. Washington said Meta remained Supergut’s most efficient advertising platform, but said there are early indications of “TikTok ads performing as well and in some cases better than some of our Meta campaigns.”

TikTok’s command of the U.S. digital-ad market more than doubled in 2022, Insider Intelligence said, thanks to its nearly 100 million U.S. monthly active users, the virality of the platform and its hold over Gen Z, millennials and influencers. Still, its overall share remains small, accounting for 2% of U.S. digital-ad spending, according to Insider Intelligence, which expects that number to grow to 2.5% this year.

The gains came as TikTok has received increased scrutiny from Washington, D.C., over national-security concerns U.S. officials say the Chinese-owned app poses—accusations that TikTok has disputed. The company also faces competition from Google’s YouTube and Meta’s Instagram, which have launched offerings that mimic its short-form video-recommendation format.

One already large digital-advertising player that gained market share in the U.S. last year was Amazon, whose ad business is powered by its ability to target users by their purchase and browsing history. The e-commerce giant accounted for 11.7% of U.S. digital-ad spending last year and is poised to grow to 12.4% in 2023, Insider said.

“Our advertising is at the point where consumers are ready to spend,” Amazon Chief Financial Officer Brian Olsavsky said on the company’s October conference call.

Sam Bloom, chief executive of digital-ad agency Camelot Strategic Marketing and Media, said some of its clients that had withheld selling on Amazon started doing so last year, using Amazon ads to promote their products on the platform.

Other retailers have followed in Amazon’s footsteps by building digital-ad businesses based on their consumer data, known as retail media networks. Combined, Walmart Inc., eBay Inc., Etsy Inc. and Instacart took in about 1.4% of digital-ad dollars spent in the U.S. last year, according to Insider.

Edwin Fu, chief executive of Placements.io Inc., a tech company that works with digital- publishers and online retailers on billing systems for their ad inventory, described the rise of retail media networks as a “massive race to do what Amazon did.”

Streaming services also are commanding a larger slice of the digital-advertising pie. Mr. Bloom said many of its clients are shifting their spending away from traditional television and toward video platforms in an effort to reach younger viewers.

Insider said Roku Inc., Walt Disney Co.’s Hulu, Paramount Global’s Pluto TV and Paramount+, Fox Corp.’s Tubi and Comcast Corp.’s Peacock accounted for about 3.6% of U.S. digital-ad spending last year. The trend is expected to accelerate now that the streaming industry’s two-largest players, Netflix Inc. and Disney+, have launched ad-supported versions.

Vincent Létang, executive vice president of global market research at Magna, a media- investment firm that is part of Interpublic Group of Cos.‘ Mediabrands, called Netflix and Disney’s entry into the market “a game-changing moment” for ad-supported streaming. “They bring a potentially huge number of viewers,” he said, and a wealth of premium video content.

Given the emergence of new digital-advertising alternatives, Google and Meta’s declining market share is hardly surprising, said digital ad consultant Ratko Vidakovic. The surprise was actually their 2021 performance, an aberration that he said was caused by the pandemic.

U.S. digital-ad spending grew by a whopping 41% in 2021, according to media and data giant GroupM, part of WPP PLC. That growth number dipped to 10.7% last year and is expected to slip to 9.2% this year. GroupM expects overall ad spending in the U.S.—excluding political advertising—to grow by 5.5% this year.

Despite the expectations of slowing growth, digital-advertising platforms will take in an ever-larger share of ad dollars this year, GroupM said: Nearly two-thirds of total U.S. ad spending is expected to go to digital advertising in 2023, compared with less than half in 2019, the last year before the pandemic.

American’s $800B investment in corporate R&D, venture-capital investment, and government funding/subsidies represents 3.8% of GDP, maintaining an edge over China’s $660B (2.7% of GDP)

“…The Economist has totted up corporate spending on R&D, venture-capital investment, direct government funding and, for advanced technologies, implicit funding through subsidies, and subtracted the overlap among these categories. This calculation confirms that America maintains a slight edge, spending about $800bn or 3.8% of GDP in 2020. That compares to about $660bn in China after adjusting for differences in the cost of living, or 2.7% of GDP….”

Economist Staff, "China and the West are in a race to foster innovation,"The Economist, October 13, 2022, https://www.economist.com/briefing/2022/10/13/china-and-the-west-are-in-a-race-to-foster-innovation

China and the West are in a race to foster innovation

China’s government is planning on winning the ai race, winning future wars and winning the future,” warned Todd Young, an American senator, in July. “At stake” in the West’s technological competition with China, echoed a report from American officials and businessmen in September, “is the future of free societies, open markets, democratic government, and a world order rooted in freedom not coercion.” This week the head of a British intelligence agency joined the chorus, urging “deep investments” in new technology to counter China’s growing prowess.

The anxiety is easy to understand. In 2008 China spent a third as much as America did on research and development (r&d) and about half as much as Europe, after adjusting for differences in the cost of living. By 2014 it had surpassed Europe. By 2020 its spending was 85% of America’s.

The fruits of this investment are becoming apparent: in August a Japanese research institute calculated that China now produces more of the world’s most highly cited academic research than America does. Since 2015 more patents have been issued in China than in America. China’s output of a basket of sophisticated goods including information technology, pharmaceuticals and electronics is expected to surpass America’s this year, according to a report published by the Information Technology and Innovation Foundation, an American think-tank. “China has become a serious competitor in the foundational technologies of the 21st century,” concluded another report last year from the Belfer Centre at Harvard University.

Small wonder, then, that Western countries are embarking on a frantic effort to retain or regain their technological edge. On October 7th America issued fierce new restrictions on exports to China of advanced semiconductors and related equipment. The new rules could be as crippling to the Chinese chip industry as previous American sanctions were to Huawei, a Chinese telecoms firm, says Greg Allen, who used to head the artificial-intelligence (ai) unit at America’s Department of Defence. “It’s a total clamp down, trying to cut off every head of the hydra of China’s chip industry.”

As well as trying to disrupt the flow of technology abroad, America’s government is investing more in innovation. In August Congress approved $370bn of spending on green energy, including lots of money for research. The month before it passed the Chips and Science Act, which provides $52bn over five years for the semiconductor industry, some of which will incentivise private r&d.

The act also revamps the National Science Foundation (nsf) to put more emphasis on applied science and technology and potentially doubles its funding. Germany, Japan and South Korea are making multi-billion-dollar investments in computer chips. Last year Britain announced the $1bn Advanced Research and Invention Agency (aria) to supercharge high-risk, high-reward science.

The result of all this is a global boom in investment in innovation. In 2020 the world’s spending on r&d exceeded $2.1trn, over 2.5% of global gdp, a record. The binge has three notable features. The first is the heavy involvement of governments, which are unwilling to leave investment to capital markets and are instead both funding r&d and subsidising production of certain high-tech goods. Both China and the West explicitly link such spending to geopolitical competition. “Technological innovation has become the main battlefield of the international strategic game,” said China’s president, Xi Jinping, in a speech last year to Chinese scientists. “We’re in a multi-generation era-defining competition against the ccp,” rhymes Mr Young, one of the sponsors of the Chips Act.

The second feature of the new era is experimentation with different types of funding that couple industrial policy with efforts to promote risk-taking or private-sector rigour. America and Britain, for instance, are reviving research missions akin to America’s cold-war quest to put a man on the Moon. China, meanwhile, is using “guidance funds”, in which the state takes a stake alongside private investors, to steer money to startups in ai and chips, among other advanced technologies.

Third, governments are trying to ensure their country captures more of the benefits of innovation. That can mean both preventing exports of some goods and using industrial policy to promote domestic production.

But there remain big differences in approach between China and the West—most notably the far more muscular role the state still plays in directing innovation in China to favoured industries. The West, in contrast, relies on a more diffuse network of universities, non-profits and private businesses that have more freedom to set their own priorities. There is little doubt that China’s system has helped it catch up with the West in some existing technologies, but analysts question whether it will be as good at generating future breakthroughs. The answer will determine the outcome of the global battle for technological dominion.

America’s government invested lavishly in innovation during the cold war, through such organisations as the nsf and the Defence Advanced Research Projects Agency (darpa). Its spending peaked at 1.86% of gdp in 1964. But after the fall of the Berlin Wall federal spending on r&d fell well below 1% of gdp. Private investment, meanwhile, doubled from 1% of gdp in 1979 to 2% in 2017. Giant tech firms such as Google, Facebook (now Meta), Amazon and Apple sprouted in America. China spawned similar titans, such as Alibaba, Baidu, JD.com and Tencent.

But on both sides of the Pacific the age of free-flowing private capital left many disappointed. The Communist Party has called the spread of big consumer-tech firms a “disorderly expansion of capital”. It has obliged China’s internet giants to follow its priorities, blocking share sales and issuing abrupt regulations to cow wayward firms. It seems to want less video-gaming and online commerce and more ai, chips and green tech.

Many Americans have similar misgivings. Peter Thiel, a fabled investor, has argued that there has been too much investment in “bits” (software and analytics) and not enough in “atoms” (hardware and manufacturing). “With chips we were caught behind the eight ball,” says Eddie Bernice Johnson, a Democrat from Texas who chairs the committee that drafted the Chips Act, “It was a national security imperative.” Mr Young of Indiana agrees: “The totally free-market theories of Friedman, Hayek—they don’t make sense when you’re facing an existential threat that plays with market forces.”

Buying breakthroughs

To allow a proper comparison of the sums devoted to innovation on the two sides of the Pacific, The Economist has totted up corporate spending on r&d, venture-capital investment, direct government funding and, for advanced technologies, implicit funding through subsidies, and subtracted the overlap among these categories. This calculation confirms that America maintains a slight edge (see chart 1), spending about $800bn or 3.8% of gdp in 2020. That compares to about $660bn in China after adjusting for differences in the cost of living, or 2.7% of gdp.

What is more, juguo tizhi can also lead to misallocation of funds. A paper published in the journal Econometrica in July suggests that Chinese spending on r&d spurs less growth in productivity than that of neighbouring Taiwan. That is in part because the state often supports soes, even if they are less productive. Several studies suggest that corporate r&d in China is about half as productive as in America (although they do not focus exclusively on advanced technology).

America’s expenditure, meanwhile, is much more diffuse. Private businesses account for about 60%, venture capital for nearly 20% and foundations, charities and universities more than 5%. In a recent presentation, Pierre Azoulay, a professor at mit, notes that the “Cambrian explosion of philanthropic funders” is a “silver lining” compared with the perceived sclerosis in government funding. From 2010 to 2019, research funding from non-profits nearly doubled, from $12bn to $22bn. “Our system is unique because its more distributed and bottom-up; not top-down,” says Maria Cantwell, another senator.

America may also be more daring in its investments. The Institute for Progress (ifp), an American think-tank, is helping government agencies distribute grants more effectively, says Caleb Watney, a co-founder. Erwin Gianchandani of the nsf cites “golden tickets” as an example. Rather than the standard consensus-based process to allocate funding, a single reviewer can champion a project.

Michael Lauer, head of extramural research at America’s National Institutes of Health (nih), lists a handful of new programs where labs receive funding with far fewer strings attached than normal. The Other Transactions Authorities, a recent nih program that quickly funds unconventional projects, disbursed over $2bn in 2020 and 2021.

Hunting for lightbulbs

America is also creating more “moonshot” programs in an effort to replicate the success of darpa. Last year it launched the $1bn Advanced Research Projects Agency for Health (arpa-h) to focus on ambitious biomedical research. The total amount of funding in this category increased from about $4bn in 2021 to nearly $6bn in 2022. Tom Kalil of Schmidt Futures, another organisation that aims to shape policy on innovation, says this heralds a shift towards more risk-taking. China lacks any equivalent funding agencies, notes Mr Cao of Nottingham University China.

It is possible to exaggerate the strengths of both the American and Chinese systems. For all the talk of moonshots, notes a former White House official, nih was still slow to fund research on covid-19 at the beginning of the pandemic. And researchers are still often buried in paperwork. The new funding does little to address the administrative burden in the current system—researchers lose about 40% of their time to that, notes Tony Mills of the American Enterprise Institute, another think-tank. By the same token, although thousands of new firms have sprung up in China in favoured industries, such as ai and semiconductors, most of them do not seem to have achieved much.

Neither system has a monopoly on results. According to a report published in September by the Special Competitive Studies Project, a research group organised by Eric Schmidt, a former ceo of Google (and a former member of the board of The Economist), China is dominant in some industries, such as 5g telecoms. It makes some 80% of the world’s lithium batteries. But the West is ahead on biotech, cloud computing and ai. It has been the source of most fundamental advances in these fields, such as crispr (a gene-editing technology) and the transformer architecture that underpins many big ai models.

Although few advanced computer chips are made in America, American firms tend to design them. asml, a Dutch manufacturer of chipmaking equipment, has a monopoly in the advanced lithography needed to make the fanciest ones. And the Chips Act is prompting Intel and tsmc, two of the biggest chipmaking firms, to build new semiconductor fabs in America. Intel is also set to spend nearly $20bn on new chip factories in Germany.

China has built first-rate ai models by throwing money at researchers and firms. Wu Dao, its version of gpt-3, an American ai model that can write like a human, uses ten times more parameters to train itself. Yet the chips used for much of this sort of machine-learning, gpus, were originally developed to generate graphics for video games—one of the industries on which the Chinese government has cracked down most ferociously. By the same token, it was not until September that a Chinese firm developed a vaccine against covid-19 that is as effective as Western ones—a recent high-stakes test of its capacity to innovate. “China does best in products where manufacturing is complex but the science is mature, for example in batteries,” says Dan Wang, an analyst at Gavekal Dragonomics, a research group.

But China is trying to mend some of the failings of its system. It boosted funding for basic research by 16% last year, in an attempt to foster more breakthrough discoveries. It is also trying to reduce centralisation. In July the party announced new rules to increase scientists’ autonomy. “There is evidence that China has recognised the limits of using a blunt metric to evaluate scientists,” adds Mr Wang. “Thus universities are starting to move towards the peer-review system of the West.”

There are some mistakes, however, from which China shows little sign of retreating. Mr Xi’s decision to rein in the tech industry contributed to an 11% contraction in venture-capital investment from the first three quarters of 2019 to the same period in 2022. In America vc investment grew by 70% in that time. China’s stubborn zero-covid policy, meanwhile, is driving foreign capital and talent out of the country. A survey by the German Chamber of Commerce in May found that nearly a third of foreign workers plan to leave.

Such talent is crucial to competitiveness. In the past China benefited from both foreign investment and large cohorts of “sea turtles”, students and researchers who work or study abroad and later return. The share of Chinese students studying abroad who eventually returned home rose from 25% in 2004 to 65% in 2019. And from 2015 to 2019, the number of academic papers published involving co-operation between American and Chinese researchers grew by over 10% a year, according to Nature.

Yet in 2020 this growth in academic collaboration between the two countries abruptly stalled. Less than half as many Chinese received visas to study abroad in the first half of 2022 as in the first half of 2019. This pulling apart is bad for the world, but China may suffer more. It does not have as diverse a pool of researchers as the West. According to data from MacroPolo, a think-tank, although 60% of the world’s best ai researchers work in America, over two-thirds of them are foreign (and over a quarter of them Chinese). In contrast, China draws overwhelmingly on domestic talent: almost all its best ai researchers are Chinese, and 70% of them have studied only in China.

America is not only open to foreign expertise, it also benefits from a big network of alliances with other technologically advanced countries. Collectively, America, Britain, France, Germany, Japan and South Korea spend over twice as much on r&d as China. China, by contrast, has few allies, and none that are powerhouses of research and innovation.

Bulbs blown

American politicians do not seem to understand the advantage conferred by their country’s openness, however. The original draft of the Chips bill included a provision to boost skilled immigration. Although some politicians, including Mr Young, gave it cautious support, it had to be removed to ensure the support of more Republicans, in particular. (America’s allies, happily, are doing better. Britain has devised a scheme to provide visas to graduates of top universities. Australia and Canada, already home to lots of immigrants, are increasing immigration further.)

Whatever the limits to America’s openness, however, China’s growing isolation—both self-imposed and enforced by restrictions like America’s new rules on tech exports—is far more severe. After several years when its technological rise seemed unstoppable, the outlook suddenly seems much less clear. In the coming days Mr Xi will preside over China’s 20th Party Congress. Across the Pacific, America’s Congress will be debating how much money to devote to the Research initiatives at agencies like the nsf. You can be sure each will be on the other’s mind.

We examine theinfluence of physical proximity on between-startup knowledge spillovers at one of the largest technology co-working hubs in the United States. Relying on the random assignment of office space to the hub's 251 startups, we find that proximity positively influences knowledge spillovers as proxied by the likelihood of adopting an upstream web technology already used by a peer startup. This effect is largest for startups within close proximity of each other and quickly decays: startups more than 20 meters apart on the same floor are indistinguishable from startups on different floors. The main driver of the effect appears to be social interactions. While startups in close proximity are most likely to participate in social co-working space events together,knowledge spillovers are greatest between startups that socialize but are dissimilar.Ultimately, startups that are embedded in environments that have neither too much nor too little diversity perform better, but only if they socialize.

Maria P. Roche, Alexander Oettl and Christian Catalini, "(Co-)Working in Close Proximity: Knowledge Spillovers and Social Interactions,"National Bureau Of Economic Research, June 2022, https://www.nber.org/papers/w30120

Core Findings

“…The setting for our study is one of the largest technology co-working spaces in the United States. The building consists of five floors, covering 9,300 m2 (100,000 sq.ft.). One challenge in examining the relationship between location and knowledge spillovers is that rms and individuals may choose to locate in areas where knowledge exchange is already likely to be high. To deal with this potential endogenous location choice, we rely on the random assignment of office space to the hub's 251 startups. We measure knowledge spillovers as the instance of adopting a component of a peer rm's technology stack. Using floor plans to measure geographic distance, we find that close physical proximity greatly influences the likelihood of these knowledge spillovers. This effect, however, quickly decays with distance where startup rms that are more than 20 meters (66 feet) away are no longer influenced by each other. Strikingly, being located more than 20 meters apart, but on the same floor does not appear to differ from being located on a different floor altogether. In addition, we find that when firms overlap with common areas at the hub (e.g., kitchens), the distance of influence increases, revealing the important role that these spatial features play in extending geographic reach and in promoting knowledge spillovers..”

The Evidence

“. .Table 3 presents the results from assessing the effect of distance on the amount of peer technology adoption (ln(AdoptCountij + 1)) using a standard OLS model and using a linear probability model to estimate the likelihood of adopting a technology from a peer rm 1(AdoptTechij ). In the full model (Columns 2 and 4), using rm-x-room fixed effects and controlling for industry, business model, gender, age and pre-period technology overlap, we find that the doubling of distance between two dyads reduces both the amount of peer technology adoption by 3.5% and the likelihood of any peer technology adoption by 1.7%, with both point estimates significant at the 1% level. As seen, the magnitude and statistical significance of the effect remains largely unchanged with the inclusion of additional controls….”

The empirical evidence presented in this article suggests that VC financing is positively associated with firm innovation and growth in the United States... Figure 6 displays the shares of patents and patents adjusted by the quality (proxied by citations) of VC-backed companies. It confirms the increasing importance of VC-backed companies for innovation...Regression analysis is now conducted to evaluate the performance of VC-backed and non-VCbacked firms along four dimensions for the years following an initial public offering (IPO) of stock: the R&D-to-sales ratio, growth rate of employment, growth rate of sales, and firm market value (in natural logarithm). The results are presented in Table 1. The regressions are based on U.S. public companies between 1970 and 2014. To compare VC-backed companies with their non-VC-backed counterparts, a VC dummy is entered as an independent variable that takes the value of 1 if the company is funded by VC before its IPO. In all regressions, industry dummies, year dummies, and a year dummy for the IPO are included. In addition, a cross term is added between the VC dummy and the number of years since the firm’s IPO. As shown in the first row of Table 1, VC-backed companies are more R&D intensive and grow faster than their non-VC-backed counterparts. On average the R&D-to-sales ratio of a public VCbacked company is higher than its non-VC-backed counterpart by 5.2 percentage points, and it grows faster—by 4.9 percentage points in terms of employment and 7.0 percentage points in terms of sales. These superior performances translate into higher market values: VC-backed companies are valued 37.3 percent higher than their non-VC-backed counterparts. The difference in performance, however, gradually dwindles over time, as shown by the negative estimates of the regression coefficients in the second row. As a consequence, the performances of VC- and non-VC-backed public companies tend to converge in the long run, though the speed of convergence is fairly low, as revealed by the magnitudes of the estimates in the second row....As demonstrated in Table 3, increasing VC funding in an industry in a given year is associated with a higher growth rate of employment and sales in the subsequent three years. According to IV regressions (2) and (4), a one-standard-deviation increase in logged industry-level VC funding is associated with increases of 1.3 percentage points and 1.9 percentage points in annual employment and sales growth, respectively, following funding.

Jeremy Greenwood, Pengfei Han, and Juan Sánchez, "Venture Capital: A Catalyst for Innovation and Growth,"Federal Reserve Bank Of St. Louis, April 2022, https://files.stlouisfed.org/files/htdocs/publications/review/2022/04/06/venture-capital-a-catalyst-for-innovation-and-growth.pdf