Business

Data shows the true productivity cost of the city exodus

Researchers have found proof of how much the end of the office will impact productivity
Image may contain Flooring Human Person Pedestrian and Floor
Getty Images / Spencer Platt

When Kodak began its slow slide towards bankruptcy from its peak in 1996, the community of scientists and inventors working at the company’s headquarters in Rochester, New York, dwindled and vanished. But their exodus caused a ripple effect: since then, the inventors in the city who were not associated with Kodak produced around 20 per cent fewer patents, and the ones they did devise were of lesser quality.

“I find it [the Kodak story] so compelling because it’s a case where the size of the high-tech cluster in Rochester changed dramatically, very quickly,” says UC Berkeley economist Enrico Moretti. “And what was clearly jumping out of the data is that the same people became less productive,” he says. “As the cluster around them shrunk, they started patenting less, and the quality of their patents was lower.”

The fallout caused by Kodak’s demise shows how brutal an exodus of talent can be for the companies that surround it, according to Moretti. His new research, published in March this year, shows how interconnected business ecosystems truly are, and how companies, through sheer geographical proximity, can impact each other’s productivity.

This matters for workers, too: people in the tech and biomedical sectors who move from a small industry cluster to a larger one like those in Silicon Valley, Seattle, or Austin become 12 to 14 per cent more productive, Moretti’s research shows. They become more productive almost immediately after they move, and they sustain their productivity for several years afterwards. People are also more likely to cite the work of other people in the same city as them, he says.

The effect Moretti documented could have several explanations. The first is knowledge spillovers between workers, which he says “captures the knowledge flows between workers within a firm, but also between firms, when people meet serendipitously and exchange ideas”. Another, which Moretti thinks is becoming increasingly important, is the quality of the talent pool which is larger and better in bigger clusters. “In a large labour market, you have a better match between workers and firms, especially when workers are very specialised and firms are also very specialised, and very different from each other,” he explains. “It makes a huge difference for most of our careers, and for wages, and for the difference between cities, because larger cities allow a better match between workers and firms, which means they can be more productive.”

Moretti’s research proves that companies can impact each other’s productivity through proximity alone. But until now, nobody has actually measured how knowledge spillovers actually affect a company's bottom line in a single location.

Using data from confidential tax returns of Canadian firms in sectors including real estate and professional services, a team of researchers has done this for the first time. University of Toronto urban economist Nathaniel Baum-Snow, Nicolas Gendron-Carrier at McGill University in Montreal and Ronni Pavan at the University of Rochester mapped out the learning and knowledge spillovers that happen between workers in the same building or adjacent buildings. By analysing the spillovers at the very local level, they could isolate the effects of learning and distinguish them from recruitment, which operates at larger scales than a city block, they argue.

The researchers discovered that when a firm with double the revenue of its neighbour moves into a city block, it increases its neighbour’s productivity by about two per cent – “not huge, but it’s definitely there in the data,” Baum-Snow says. The benefit is bigger for more productive firms. What matters is the average quality of the information you get from companies close by, rather than total number of neighbours (and their total revenue) on the block as a whole, Baum-Snow explains. Although they can’t infer the details of these interactions from the firms’ tax data, Baum-Snow believes gatherings like formal work meetings, informal coffees, or gatherings at the pub would contribute to this effect.

Baum-Snow’s team found that knowledge transfers may be more valuable within occupations but across industries. For example, a data scientist at a tech company seems to benefit more from interacting with a data scientist at an investment firm, than one at another tech company. “Part of what is going on is direct vs indirect search for information,” Baum-Snow says. “So randomly running into your neighbour and hearing about these things is more likely with simple proximity.”

But the pandemic city exodus is causing these spillovers to dry up. Although there is not much evidence for a drop in productivity so far, Moretti thinks that in the long run, there will be a loss of innovation and productivity as a result of the pandemic shift to remote work. “It’s hard to create new ideas if you’re physically isolated. It’s hard to interact with clients and ask them exactly what new things they need. Also, it’s hard to onboard new people into teams,” he says. 

For this reason, he thinks most workplaces won’t remain 100 per cent remote after the pandemic is resolved. “There will be a number of occupations that can be done fully remotely [like computer coding or administrative work]. But I don’t think that’s the majority,” he argues.

That may explain why some workers are already moving back to places such as New York and San Francisco despite the high rates of infection. Nevertheless, there is a disconnect between what managers want and what people working for them want. Although many managers would prefer to see employees in the office three to four days a week, most employees would like to be in the office one to two days per week, according to surveys by research company Global Workplace Analytics. Kate Lister, who is president of the company, says businesses’ top consideration right now should be attracting and retaining talent, and they will lose the best and brightest if they force people back to the office full-time. She argues that hiring remote workers expands the talent pool to other parts of the country and the world.

Thomas Cornelissen, an economist at the University of Essex who has studied spillovers among employees, thinks it is possible to have it all. “With the hybrid model, workers could realise the best of both worlds," he says, "having some days in the office for the positive spillovers, and having some days where people benefit from working from home as well.”

More great stories from WIRED

This article was originally published by WIRED UK