The Benefits of High-Tech Job Growth Don't Trickle Down

Cities around the United States and the world see high-tech industries as a key source of economic advantage: Look no further than the incredible time, expense, and effort more than 200 cities put into competing for Amazon’s HQ2. Urban leaders know that, not only are jobs in technology good jobs, they also spur the creation of lots of follow-on jobs. In The New Geography of Jobs, economist Enrico Moretti argues that every high-tech job has a multiplier effect of an additional five jobs, spurring a virtuous cycle of economic development and employment growth.

Yet it is increasingly evident that major tech hubs (San Francisco, Boston, Seattle) and superstar cities with large high-tech industries like New York and London have become more and more unequal, especially with their high housing costs eating away at the incomes of blue-collar and service workers.

Now, a new study finds clear evidence that low-skilled workers fail to benefit from high-tech growth and development. Even in the face of very strong job-creating multipliers, low-skill service work pays so poorly that the average wage for these workers actually falls, with their income further eroded by high housing costs.

In a forthcoming article in the journal Research Policy, Neil Lee of the London School of Economics and Stephen Clarke of the Resolution Foundation in London, track the growth and effects of high-tech and digital industries across cities in the United Kingdom from 2009 to 2015. (The high-tech and digital industries combined, account for roughly 7 percent of total U.K. employment.) Lee and Clarke look at the effect of these industries across nearly 200 U.K. commuting zones—smaller geographic areas where people regularly travel for work. Their study examines both the multiplier effects of these industries and effects on the wages and quality of jobs in less-skilled industries such as construction, retail, hospitality, and restaurants.

First, Lee and Clarke find the jobs multiplier to be considerably smaller than in previous studies focusing on U.S. high tech. They find that 10 new high-tech jobs create roughly seven additional jobs, of which six are very low-wage service jobs. This represents a multiplier of less than one to one (with every high-tech job creating 0.7 follow-on jobs). The researchers’ models generate a range of multipliers, from 0.4 on the low end to 1.8 to 2 .0 on the high end. But whatever the case, the figures they generate are considerably lower than the four-to-five-job multiplier effect identified in the U.S. by Moretti. This held true even when the researchers used the same economic-multiplier models as Moretti.

The study’s estimates are more in line with other research on high-tech industries in Europe, which identify a multiplier of a little better than one to one (about 1.1 follow-on jobs for every new high-tech job).

Lee and Clarke note several reasons why these high-tech multipliers may be lower in the U.K. than in the U.S. The British definition of service jobs may be narrower, for one thing, and U.K. commuting zones tend to be smaller than U.S. metropolitan areas, meaning that multiplier jobs are possibly being created outside them. Or it could speak to a difference in how places either side of the Atlantic responded to the economic crisis. Migration is also higher in the U.S., which may affect the multipliers.

Second, the study finds that wage gains from high-tech growth are very uneven. High-tech growth leads to better jobs and higher wages for more skilled workers. But it leads to lower wages for less-skilled workers. These effects are compounded by housing costs, with less-skilled workers being even worse off when housing costs are taken into account.

Indeed, the researchers see “a negative and statistically significant effect from high-tech on wages for workers in the bottom third of educational attainment.” This effect is even larger when local housing costs are included, which stands in sharp contrast to the situation for higher-skill workers: Their effective wages rise when controlling for housing costs.

The reason for this unevenness boils down to the fact that high-skilled tech workers are mobile and paid at rates that factor in higher housing costs, whereas less-skilled workers are more or less trapped; people competing for low-wage jobs mostly lack the resources to move to other places.

The study adds valuable new evidence to our evolving understanding of high-tech development and inequality. Attracting high-tech growth—say by going after HQ2s, or even by stimulating local high-tech clusters and startup ecosystems—is a limited form of trickle-down economic development policy, one based on faulty assumptions. High-tech growth appears to have much weaker multipliers than previously thought. And it offers no panacea for low-wage jobs: If anything, it makes a bad and highly unequal situation even worse.

It doesn’t make sense to throw the proverbial baby out with the bath water and neglect (or worse, try to limit) high-tech development. High-tech industries are crucial motors of innovation and productivity that power the economy broadly. The challenge is to forge economic-development strategies that enable more workers to gain, by preparing less-skilled workers for better jobs and upgrading low-wage, precarious service jobs into higher-paying and more knowledge-based work.