Do large firms generate positive productivity effects?

Numerous studies have documented the growing dominance of large firms over the past few decades in many industrialized countries. Many research articles focus on the potential negative effects of this increased market concentration, which raises concerns about market power in both labor and product markets. In a new study, we investigate whether large firms also generate positive effects. Our research shows that large firms generate significant positive effects on total factor productivity (TFP) to their local suppliers. To date, these types of side effects have only been identified for multinational enterprises located in developing countries. Using inter-firm transaction data for an industrialized country, Belgium, we find that large local firms as well as multinationals generate positive TFP spillovers.

Event research

We use the universe of firm-to-firm transaction data for Belgian firms between 2002 and 2014 to investigate whether a firm located in Belgium that starts a new relationship with a superstar firm has higher TFP after the start of the relationship . (TFP reflects technological and organizational changes that increase output for a given quantity and quality of inputs.) We define a firm as treated if it reaches a point where more than 10 percent of its sales are to a superstar firm, for three different types of superstars: large firms, multinationals and exporters. The TFP of treatment firms is compared to the TFP of a control group comprising firms that never sell to a superstar firm.

The first chart below plots the TFP of a “treated” firm for each year before and after the treatment, for example with a “1” on the horizontal axis indicating the year of treatment and all points showing the effect relative to the year before treatment (“0”). The chart shows that up to three years after the event, firms located in Belgium that started selling to a superstar firm enjoyed about 7 to 8 percent higher TFP than the control group. Interestingly, the size of the spread is about the same for all three types of superstar firms. This result suggests that the effects do not emerge from the partner firm being multinational per se, but rather from the superstar firm being more productive and successful. These are not the same. We show that these efficiency effects exist even if a large firm is not multinational or exporting.

Creating new relationships with a superstar firm boosts productivity

Liberty Street Economics three-panel graph comparing the first time a firm begins a major supply relationship with a superstar firm versus a control group of firms that never sell to a superstar firm.  In the diagram, a superstar firm is defined as a multinational (left panel), a large exporter (center panel), and a very large firm (right panel).

Source: Amiti et al. (2023).
Notes: These results are derived from event studies comparing the performance of firms that initiate major supply relationships with a superstar firm (the treatment group) with firms that do not initiate such relationships (the control group). In the left panel, the superstar is defined as a very large firm (top 0.1 percent of the sales distribution), in the center panel, the firm is multinational, and in the right panel, an exporter. The y-axis is on a logarithmic scale, so 0.5 = 5 percent. N is the number of observations. CI is a confidence interval.

Of course, very large firms are also often multinational. But we show that the prevalence of superstars is present even when we consider the initiation of relationships with very large firms that are no multinational companies in the graph below. Furthermore, we show that there is no effect of forming strong supply relationships with small firms, suggesting that the superstar relationship is causal.

Positive productivity growth implies that a firm must also grow in scale, and indeed we also see sales jump by about 28 percent for supplier firms. This effect remains even after netting out sales going to the superstar firm. Likewise, we see large increases in intermediate inputs, labor and capital, and the number of non-superstar buyers.

Positive productivity effects for new suppliers even if the large firm is not multinational

Liberty Street Economics two-panel graph chart showing the spread of total factor productivity for a firm that begins a large supply relationship with a superstar firm that is large but not a multinational enterprise (left panel) and one that is large and is also multinational enterprise (right panel).

Source: Amiti et al. (2023).
Notes: These are the same event studies as the first panel of the chart above, but we divide very large firms into those that are multinational (right panel) and those that are not (left panel). The y-axis is on a logarithmic scale, so 0.5 = 5 percent. CI is a confidence interval.

What are the mechanisms behind the proliferation of superstars?

The classic reason for diffusion is the transfer of know-how. We show that superstars that have higher R&D, more managerial know-how/skills, and are more IT-intensive generate the largest spillover effects. The analysis also finds that new superstar providers have higher overall profits, but the average markup falls because the superstars will capture a portion of the relationship rents. While the supplier has a lower markup on its sales to Superstar, it increases its overall profits by expanding the number of buyers it delivers to, both within and outside of Superstar’s network.

We also show new evidence of non-performance spillovers generated when a relationship with a superstar firm helps a supplier gain access to a new network of potential customers. We call this the “dating agency” effect to reflect the role of the superstar firm in finding partners. This additional benefit can only work through reducing the costs of searching for suitable buyers or through a signaling effect where dealing with the superstar firm causes other firms to update their beliefs about the supplier’s quality (and these signaling effects are particularly strong in a -network). Indeed, we find particularly large positive effects on the number of buyers in the superstar’s network, consistent with the dating agency effect.


Governments spend large sums of money to attract and retain multinational companies, in part because of their belief in the importance of these supply chain benefits. Our results highlight that being global per se is not necessary to generate side effects. We show that large domestic firms generate TFP spillovers of the same magnitude as multinationals. Although there may be potential costs associated with the dominance of large firms in the modern economy (identified, for example, in studies of market power and political influence), our work shows some advantages in allowing superstar firms to grow and create relationships with less successful firms.

Mary Amity is head of Labor and Product Market Studies in the Research and Statistics Group of the Federal Reserve Bank of New York.

Cédric Dupre is an economist at the National Bank of Belgium.

Josef Konings is Professor of Economics and Dean of the Nazarbayev University Graduate School of Business and Director of the Center for Regional Economics (VIVES) at KU Leuven.

John Van Reenen is the Ronald Coase Chair in Economics and a lecturer at the London School of Economics.

How to cite this post:
Mary Amity, Cedric Dupre, Joseph Konings, and John Van Reenen, “Do Large Firms Generate Positive Productivity Effects?”, Federal Reserve Bank of New York Freedom Street Economics12 Oct. 2023,

The views expressed in this publication are those of the author(s) and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the author(s).

Leave a Comment

Your email address will not be published. Required fields are marked *