Coachella diaries – part 1

The OECD’s Global Productivity Forum is like Coachella for economists (if all the acts at Coachella lacked diversity and all the concerts were played in the dated function room of a downtown Marriott). This year, the Forum took place in Ottawa, co-hosted by the Bank of Canada and Innovation, Science and Economic Development, Canada. The main stage saw performances from the likes of Dirk Pilat, Chiara Crisculo and James Bessen, as well as and local acts Kevin Fox and Dan Andrews.

But as with any good music festival, it was on the smaller stages that you find the gems.

The highlight from day 1 was Antonin Bergeaud’s presentation on the relationship between innovation and inequality. Using an incredibly powerful data set, Bergeaud is able to pinpoint the impact innovation (measured in terms of R&D expenditure per employee) has on wages – at the firm level.

It’s probably of no surprise that the study finds wages are higher in innovative firms. For every 1 per cent increase in R&D intensity, the average wage paid to workers was shown to increase by 1.6 per cent. This innovation premium survives even when accounting for differences in geography, education and tenure.

What might be more surprising is the impact that innovation has on wages paid to low-skilled workers.

The usual story we hear is that innovation leads to greater inequality: skill-biased technical change benefit those at the top of the distribution the most. But Bergeaud and his co-authors found that low-skilled workers earned “substantially more” whenthey worked in a firm with a higher R&D intensity. Moreover, they also found that increases in a firm’s innovativeness has the greatest impact on the wages of the lowest skilled.

Bergeaud offers an interpretation for why this might be the case. He argues that within the firm, the work performed by high and low-skilled employees is complementary. To ensure that high-skilled workers can concentrate on the most difficult tasks, the firm needs to identify and retain low-skilled workers with high-abilities, and steer away from any workers that aren’t a good fit. Consistent with this, the data also showed that low-skilled workers in highly innovative firms typically have much higher tenure as well.

So while yes, high-skilled workers benefit from technological advances, low-skilled workers in innovative firms do better than other low-skilled workers. This result is likely to be strongest in those sectors that are the competitive.

Just as impressive as the key findings are the instruments used to find them. The study was made possible using incredibly powerful linked employer-employee data (LEED) from the UK. You can think of a LEED as a big version of BLADE, but with individual TFNs added in. For this paper, the researchers were able to make use of a decade of ASHE data that pertained to some 340,000 individuals, living in 230 regions and working in 160,000 enterprises. (The ABS are working on constructing a LEED with Australian data, but it’s still a ways off.)

Bergeaud’s results are not necessarily inconsistent with the conventional understanding of innovation and inequality, but there’s a very important nuance that this paper helps uncover.