New paper: Innovative Events

Anna and I have a new paper out in Research Policy. Publishing this has been an epic experience: six years in the making.

The TLDR; we introduce a brand-new measure of innovative activity — media-reported product and service launches — and run it through a series of key use cases.

Why is this important? We know innovation is important for economic growth. We also know it’s very hard to measure, and existing metrics — patents, trademarks, surveys — are all pretty limited.

A handful of studies have got around these limits by working with product-level data — typically in niche cases like pharma, where new products are regulated and very clearly defined.

We advance on this by building on Growth Intelligence data, which monitors ~3,000 online media sources across all industries. We do *a lot* of work in the paper to clean, rebuild and validate this info.

Our new metrics — reported launches of new products and services — cover all UK sectors and firm types. In the paper, we focus on SMEs to keep ascription as clean as possible.

We then do three things.

First, we show that these innovative events complement patents, trademarks and surveys — with higher frequency, wider industry and geographic coverage.

Next, we show that at the firm level, past IP predicts current launch activity. That is, if you filed patents or trademarks, that makes you more likely to launch a product down the line.

Finally, we show that launch activity helps explain variations in firm performance. A minority of firms with a high-profile launches drives this. That fits existing evidence on high-growth firms: a few stars drive overall growth.

Here’s the journal version again. You can also access the pre-print here, and play around with some of the data and code.

Last word: big thanks to Alex Coad and the reviewers at RP. Alex was a stern but kindly editor, whose suggestions made the paper a lot better.