On the day the Didi deal was announced earlier this month, Uber board member Bill Gurley said Uber’s rivals in other markets had a slim chance of splitting the market with the dominant player, just as Uber struggled to erode Didi’s share in China.

After China, the industry will consolidate in other markets, said Hans Tung, an Asia-focused investor and managing partner at GGV Capital, which backed Didi and Grab, a Singapore-based ride service.

“There will be a dominant No. 1,” he said that same day.

The consensus of 11 economists interviewed by Reuters, however, suggests an entirely different scenario, one of perpetual competition in a business with relatively few barriers to entry.

“That one firm wins is a narrow and not accurate way to think about these firms,” said David Evans, chairman of the Global Economics Group and co-author of a recent book that included Uber, Matchmakers: The New Economics of Multisided Platforms.

Ten other economists who have studied ride-hailing agreed that the growing industry, which UBS estimates to be a $40 billion market, has room for at least two successful players, and perhaps a few smaller ones.

Get Data Sheet, Fortune’s technology newsletter.

The industry, they said, has none of the elements that traditionally have enabled single companies to control a sector. If it is the first of its kind, a company can dominate markets that have huge infrastructure costs, such as putting up cell towers or laying pipes; a large workforce of employees with specialized skills; and customers who get locked into a service and have difficulty leaving for competitors.

Ride services, by contrast, are relatively cheap to start, depend on contract labor with no inherent loyalty or specialized skills, and have free apps that can be downloaded in seconds.

“You may not want to try a new social networking site if your friends aren’t on it,” Evans said. “But you don’t care what app your friends use for ride-hailing.”

‘Winner-Take-All’ or ‘Room for Two’?

The question of whether on-demand ride services will remain open to new players has vexed startups and investors since Uber started the industry seven years ago.

Companies taking on Uber include Lyft in the United States, Grab in Southeast Asia, Ola in India, and newer startups like New York City’s Juno. In the United States, in particular, part of Uber’s attraction to investors is the chance at grabbing the entire industry.

In a statement, Uber said: “The ridesharing industry around the world is highly competitive and innovative. That’s good for riders.”

Uber investor and board member Gurley argued that any competitor would need to pursue a different strategy—perhaps offering more luxury and high-end services—to successfully battle Uber in its strongest markets.

Didi, Ola, and Grab did not respond to requests for comment.

When business magnate Carl Icahn invested $100 million into Lyft in early 2015, he told media outlets he saw “room for two.” Chris Sacca, a prominent venture capitalist who invested in Uber, responded, “This is a winner-take-all game,” on Bloomberg television.

Lyft has hired an M&A firm and recently explored the possibility of acquisitions by several companies, a source familiar with the discussions said, and reports of a possible sale stimulated talk of whether it could compete with Uber.

Lyft says it can. In the United States, it says it more than tripled its drivers to about 315,000 in the last year. Between October and May it nearly doubled its annual gross revenue to $1.9 billion—although that figure does not reflect the many rider discounts and promotions Lyft offers.

Uber has 1.5 million drivers and projected $26 billion in gross revenue globally this year, based on a 2015 presentation for investors.

The Magic 3-Minute Mark

Last year, Lyft hit another benchmark: The wait time for a ride is three minutes, on par with Uber, said president and co-founder John Zimmer. At three minutes or less, a passenger will almost always complete the ride.

“You need a certain level of scale to get to three minutes,” Zimmer said, referring to the number of drivers and passengers. “Once you reach that, if someone else has more scale, it doesn’t matter.”

New York-based Juno has brought on 12,000 drivers since launching earlier this year and already has hit the three-minute wait time in Manhattan, said Co-founder and CEO Talmon Marco.

“This is a fairly local industry,” Marco said. “You can be a hero in New York and you can be zero in California, and it’s OK.”

For more on the Uber-Didi Chuxing deal, watch:

In India, Uber and Ola are neck and neck around 45% of the market each after Uber’s market share fell and Ola’s rose in 2015, according to market research firm 7Park Data.

The challenge for new startups, however, is that leading companies subsidize their drivers and passengers as they prioritize gaining market share over profit. Both Uber and Lyft have spent heavily on driver bonuses and rider discounts and promotional credits.

“Everything that has happened in this space is completely artificial and funded by a glut of VC money,” said Daniel Ramot, CEO and co-founder of startup Via, which completes about 200,000 rides each week in New York.

Economists argue that Lyft can be a profitable company with roughly 20% of a market, which would allow it to reduce expenses through economies of scale. Lyft and Uber only release market share statistics selectively, but Lyft maintains it has more than a 20% share in the majority of its top 20 regions.

An electric company, by comparison, would need massive scale to achieve enough efficiency to allow for profits, said Stephen Margolis, an economist and anti-trust expert at North Carolina State University.

Max Wolff, an economist at Manhattan Venture Partners, believes competition will thrive mainly because ride-hailing technologies are not overly complicated and drivers aren’t earning enough money to be loyal to a single company.

There is room for other players even if Uber is dominant, he said: “They’re not as big, but they’re there too. They’re not some wheezing, dying remnant.”