Industrial policy is back on the agenda across advanced economies, with a growing channelling support towards large incumbent firms on the premise that they are the most capable innovators. This column uses data on French manufacturing firms to argue that this premise deserves scrutiny. Firms become large primarily through occasional, large ‘innovation bursts’ rather than by innovating at persistently higher rates. The arrival of these bursts involves an element of chance, so a firm’s current size says little about how much it will innovate in the future. Policies that entrench the position of incumbents may therefore slow down the churn that sustains aggregate growth.
Productivity growth has been persistently weak across advanced economies for nearly two decades. In response, governments have rediscovered industrial policy, with a growing emphasis on helping domestic firms achieve greater scale. The view is particularly prominent in Europe: the Draghi Report (Draghi 2024) report argued that fragmentation leaves European firms at a disadvantage in strategic sectors and called for consolidation and a competition framework that places greater weight on firms’ capacity to innovate at scale – ideas that have since shaped the European Commission’s Competitiveness Compass. Previous work has further established that industrial policy is often scale-biased in practice, even when unintended (Van Reenen 2012, Criscuolo et al. 2019, 2022).
These policies often assume that large firms are large because they are good at innovating, so that supporting them means supporting innovation. In firm dynamics theories of growth through creative destruction, following Klette and Kortum (2004) and Lentz and Mortensen (2008), firms grow by repeatedly out-innovating competitors, so size reflects accumulated innovative success.
In a recent paper (Berlingieri et al. 2026), we present findings that challenge that point of view. To do so, we use data from an annual survey that covers roughly 90% of French manufacturing output at the level of some 4,000 ten-digit product categories, which we merge with balance sheet data on the universe of French firms. This allows us to observe which products firms add and lose each year, and therefore to directly measure the process of product creation, product destruction, and subsequent (firm) growth in the data. The patterns we find sit uneasily with the conventional view of how firms grow.
Product innovation comes in bursts
The main novel pattern that the data show is that product innovation is extraordinarily concentrated. In a typical year, two-thirds of firms add no new products; most of the rest add one or two. But a small set of firms adds tens of products at once: the distribution of new products has a thick, Pareto-like tail, such that 76% of all product innovation comes from just 5% of firms. We label these episodes ‘innovation bursts.’
Figure 1 Distribution of the number of new products across firms
Notes: The figure plots the share of firms adding at least n new products against n, on logarithmic axes. The near-linear relationship indicates a Pareto-like distribution
Innovation bursts seem to come from a single innovation that spawns a large number of commercial applications. Take the example of Laboratoire Science & Nature (LSN), a French producer of natural cosmetics. LSN made only a handful of products for decades until it developed a way to use mineral-rich seawater in anti-ageing skincare in the mid-2010s. Its product range expanded several-fold in a short span of time, alongside employment and revenue. Similarly, Picture Organic Clothing developed a way to produce outerwear from recycled polyester and bio-sourced polymers and registered over ten new product designs in a single year.
The experiences of these firms are part of a broader pattern: innovation bursts enable firms to rapidly expand their product portfolio, sales, and employment, even if they are initially small. Even among firms that initially produce a single product, some add tens of products in a single year – and of these firms, all product innovation comes from just 7% of them. Robustness checks confirm that bursts reflect genuine expansions rather than reporting artifacts or acquisitions: revenue per product does not fall with the number of new products, and the concentration of product creation is robust to excluding firms that have engaged in M&A.
Rare innovation bursts drive firm size
Innovation bursts have a key implication for what separates large firms from small ones: if firms mainly expand through rare, outsized bursts, then differences in firm size may reflect the luck of drawing one of those bursts rather than persistent differences in the ability to innovate.
In standard theories of firm dynamics, firms grow through the gradual accumulation of incremental innovations, while the forces of creative destruction prevent firms from pulling too far away. Under this view, firms can only reach the top end of the size distribution by innovating at persistently higher rates than their competitors (Luttmer 2010), implying that the largest firms should be the economy’s innovation champions. In reality, this is not the case. Because firms grow through bursts of innovation, a single burst may propel a firm to the top of the size distribution. That a firm is large is thus evidence of a history of large innovation bursts, rather than a signal that it will be an innovation champion going forward.
The data support this interpretation. If large firms were systematically better innovators, the average rate of product innovation – the number of new products firms produce per existing product – should rise with size. It does not. Across the firm-size distribution, the innovation rate is remarkably flat, ranging between 5.6% and 8.3%, a fact known in the literature as Gibrat’s Law. What differs across firms is not just the propensity to innovate, but the occasional arrival of a burst.
Creative destruction matters more than the standard accounting suggests
Our data also yield new evidence on the importance of creative destruction for economic growth. We find that creative destruction – the entry and exit of products from firms’ portfolios – explains over 70% of the variation in firm growth at a one-year horizon and nearly 90% at five years. This is particularly true in the tails of the growth distribution, which account for a substantial share of aggregate revenue creation and destruction. In other words, firms primarily grow and shrink by gaining and losing products, rather than by improving the products they already produce.
The data also point to an indirect channel through which product innovation raises growth. As shown for US firms by Argente et al. (2024), revenue growth on a given product declines steadily over its life cycle, which suggests that opportunities to improve existing products dry up as they age. This means that standard growth accounting understates the importance of creative destruction, attributing to it only 15% of aggregate productivity growth in France, in line with the findings of Garcia-Macia et al. (2019) for the US (see also Klenow 2025). However, it misses the fact that, without the introduction of new products, process innovation on an ever-ageing product mix would gradually run out of steam.
What this means for industrial policy
Together, these findings complicate the case for incumbent-oriented industrial policy in three ways.
First, innovation bursts mean that firm size is a poor targeting signal. If innovation rates rose with size, directing R&D support towards large firms would concentrate resources where they are most productive. In our data, innovation rates are flat in size, and the concentration of production reflects the bursty arrival of past ideas. Policies that direct support towards large firms therefore risk rewarding past luck rather than future innovation.
Second, because most of the variance of firm growth runs through the creation and destruction of products – reinforced by the boost that new products give to subsequent innovation – the case for public support is a case for supporting product innovation and the churn it entails, by entrants and incumbents alike, rather than the position of particular firms. Policies that shield the existing incumbents from the displacement of their products slow down the process through which most growth occurs.
Third, and cutting the other way, our results caution against reading concentration itself as a sign of weak competition. In the model, a higher rate of creative destruction raises concentration, and does so around 30% more strongly than in a conventional model when measured by the revenue share of the top 10% of firms. The presence of large firms can be a symptom of a dynamic economy. The conclusion is not that large firms should be penalised, but that neither their size nor their incumbency is, by itself, a reason to direct public support in their direction.
Discussion
This column is not an evaluation of any specific industrial policy programme, and our discussion has abstracted from any strategic and security motives behind much of the current debate. Our results also do not imply that scale is never efficient: some technologies – notably intangible-intensive ones with high fixed and low marginal costs – feature genuine returns to scale and ignoring this would be costly (De Ridder 2019, 2024, Lashkari et al. 2024). Our findings do suggest that policymakers therefore face a trade-off between accommodating such scale effects, and entrenching incumbents whose size reflects the luck of past innovation bursts. An industrial policy that shields incumbents from that displacement risks slowing the growth it aims to promote.
Source : VOXeu
