From Factory Productivity to Local Economic Expansion

The firm-level evidence from Viet Nam's robot-exposed districts supplies the missing link between automation and the spatial outcomes already observed. This chapter follows the transition from robot adoption to value added, productivity, trade, employment, and local demand. It also distinguishes what registered firms reveal from what worker-level data reveal, so the reader can see why production can expand enough to offset direct labor savings without producing an equally shared improvement in work.

The question is not whether firms that adopt robots are productive. More productive firms may be precisely the firms most able to automate, export, and attract workers. The question is whether the study's exposure comparison is accompanied by the changes that a scale mechanism predicts: more output, higher value per worker, more trade, and additional employment in or around the expanding production system.

Firm expansion is visible in value added, productivity, and trade

The Viet Nam Enterprise Survey follows registered enterprises and records employment, wages, revenue, value added, capital, imports, and exports. It covers all registered firms with 30 or more employees and a representative sample of smaller registered firms, but it does not cover unregistered informal firms, family businesses, or self-employment. The enterprise evidence therefore observes the formal production system that can report output and trade; it does not observe every activity through which households absorb labor-market adjustment.

Because the trade variables are available for 2011 and 2019, the firm analysis studies changes across those years rather than the full annual sequence used for the main labor-market results. That timing matters. The enterprise estimates describe how registered-firm outcomes changed over a period that includes the robot-adoption surge, but they should not be presented as annual responses or as a direct observation of every factory's installation decision.

The pattern is consistent with scale expansion. In preferred specifications, an additional robot per thousand workers is associated with a 2.6–2.9% increase in district-level firm value added, a 0.9–1.0% increase in firm employment, and a 1.0–1.5% increase in value added per worker. The last measure is important: the result is not merely that firms hired more people. Value produced per worker also rose, which is consistent with technology raising productivity while the production system expanded.

Trade changes in the same direction. The reported exposure is associated with an 8% increase in exports and a 14% increase in imports. More imports can reflect greater use of foreign inputs, machinery, or components in an export-oriented production chain. More exports indicate that the additional productive capacity has access to buyers beyond the local market. Together, the two flows are consistent with Viet Nam's firms becoming more deeply integrated into a production system that crosses national borders.

The firm results therefore provide a concrete test of the mechanism developed earlier. If automation only reduced labor requirements while leaving output fixed, one would expect labor productivity to rise but would have less reason to expect simultaneous growth in value added, exports, and employment.