In recent years, researchers and the general public have debated whether industrial robots can bring production back home. The argument is the following: If robots can automate tasks previously offshored to low-wage countries, firms in advanced economies may have less reason to produce abroad. They will simply install robots in the home country and in this way avoid the challenges of complex global supply chains, transportation costs and geopolitical uncertainty. This is sometimes called reshoring.

But there is another possibility. Robots can make firms more productive and allow them to expand production. A larger and more productive firm may actually source more from abroad, not less.

In our recent paper, we investigate which of these two forces dominates by studying robot adoption and offshoring among Danish firms between 1995 and 2022. Using detailed firm, worker and customs data, we find that robot adoption is associated with more offshoring, not less. Firms adopting robots become more likely to offshore and increase the value of goods produced abroad.

Rather than replacing global production networks, automation appears to go hand in hand with their expansion.

Robots and offshoring: substitutes or complements?

At first glance, robots and offshoring might seem like substitutes.

Firms have traditionally offshored some production activities because labor costs are lower abroad. Robots change this calculation because they allow firms to automate tasks at home. If the cost advantage of producing abroad becomes smaller, firms may decide to reshore production.

There is, however, a competing mechanism. Automation can increase productivity and scale. As firms become larger and more productive, they may be better able to absorb the fixed costs involved in international sourcing and to coordinate more complex supply chains. Automation and offshoring could therefore become complements rather than substitutes.

Previous studies have produced mixed evidence on which mechanism dominates. One reason is measurement: much of the existing research measures exposure to robots at the industry or regional level. This makes it difficult to determine whether the firms actually adopting robots are also the ones changing their international production strategies. Our research instead looks directly at individual firms.

Using comprehensive microdata from Denmark

Denmark provides a particularly useful setting for studying this question. We combine administrative data covering firms and their employees with highly detailed customs records over the period 1995–2022.

We identify robot adoption from firms' imports of industrial robots. For offshoring, rather than simply measuring imports, we use a relatively narrow definition: a firm is classified as offshoring when it both imports and exports goods belonging to the same detailed product category. This helps us identify the international fragmentation of production rather than trade in general.

Our final dataset contains more than 440,000 firm-year observations covering approximately 48,000 firms.

Robot adopters offshore more

The central result is surprisingly clear: robot adoption is positively associated with offshoring.

After accounting for differences in firm characteristics, robot-adopting firms have an 8 percentage point higher probability of offshoring. Among firms that offshore, robot adoption is also associated with an approximately 18% higher value of offshored goods.

This is not simply because firms that already offshore are more likely to buy robots. When we account for firms' previous offshoring activity, the relationship becomes smaller but remains positive. It also remains positive when we focus on firms that had already offshored before adopting robots.

The timing tells a similar story. Before robot adoption, we do not observe significant differences in offshoring trends between adopters and the comparison group. Following adoption, however, firms become more likely to offshore, and the increase persists for several years. The value of offshored goods also increases following adoption.

This makes it difficult to reconcile the Danish evidence with a simple story in which robots systematically replace production abroad with production at home.

Automation goes hand in hand with firm expansion

Why might buying robots lead firms to become more internationally integrated?

An important clue comes from what happens inside adopting firms. Robot adoption coincides with increases in sales, employment and labor productivity. Firms also shift their workforce toward non-routine tasks. This points toward a scale mechanism.

Imagine a firm introducing robots into one part of its production process. Automation makes that operation more productive, allowing the company to produce and sell more.At a larger scale, it also becomes convenient to move additional stages of its own production abroad. Some of these can be sourced most efficiently abroad.

The result can therefore be both more automation at home and more sourcing internationally.

Not just more offshoring, but broader supply chains

Perhaps the most interesting finding is what happens when we look inside firms' global supply chains.

Robot adopters do not simply offshore more of the same products to the same foreign locations. They expand the number of products they offshore and the number of countries from which they source. The expansion is particularly pronounced for new products. Robot adoption is associated with an increase in the probability of starting to offshore a new product, while the value of offshoring of newly introduced products increases substantially more than that of existing products.

Interestingly, firms do not simply shift sourcing away from lower-income countries. Robot adoption is associated with increased offshoring to both advanced and developing economies. Firms are also more likely to source simultaneously from both groups of countries. In other words, automation appears to be associated with broader and more diversified global production networks.

Robots may change globalization rather than reverse it

Discussions about automation and globalization are often framed around reshoring: if robots become sufficiently capable and affordable, will factories simply return to high-wage economies?

Our findings suggest that this framing may be too narrow. For the Danish firms in our study, robot adoption does not coincide with retreat from international production. Instead, automation goes hand in hand with firm expansion and deeper integration into global value chains. Firms automate some activities while simultaneously expanding the range of products and locations involved in their international sourcing.

This does not mean that robot adoption can never lead to reshoring. Our results come from Denmark, a small, high-income and highly open economy with deep integration into international value chains. The findings are therefore most directly relevant to economies with similar characteristics.

Still, robots at home and global sourcing abroad can be two parts of the same strategy.

Full paper

Yan Hu, Katja Mann and Dario Pozzoli (2026), “Where Production Meets Automation: Robots and the International Geography of Production.”