Who Gets Through the Procurement Door?

Government procurement does not begin when a contract is signed. It begins earlier, in the set of firms able to discover the opportunity, finance an application, satisfy the buyer's requirements, and deliver at the necessary scale. This chapter focuses on that selection process so the reader can distinguish a firm's productive capacity from its financial capacity—and see why changing the access rules changes both the financing path and the quality of public supply.

Procurement participants are already selected before the contract

The Spanish data make the selection problem visible before any procurement shock is measured. Firms that participate in government procurement are much larger and older on average than firms outside the procurement market. The average procurement firm has about 73.5 employees, compared with 12.8 among other firms; its sales average €8.9 million rather than roughly €1.2 million; and its age averages 20 years rather than 11. These are large differences, but the distributions overlap. Government suppliers are not a separate species of firm that can be identified by a single size cutoff.

The comparison also has an important timing dimension. Among firms that later obtain a procurement contract, the ex-ante employment premium is 115 log points, and the marginal revenue product of capital—the additional value generated by capital at the margin—is 15 log points. The first measure says that future suppliers are already much larger than firms that do not win. The second suggests that their capital is especially productive relative to its cost or use. A high employment premium alone might indicate scale; the MRPK premium points toward a combination of productive opportunity and constraints that prevent firms from holding their preferred amount of capital.

That combination is why a simple winner-versus-nonwinner comparison can mislead. Suppose the firms that win public projects were already better managed, better connected, more productive, or better capitalized. Their later credit or sales growth might then reflect those pre-existing traits rather than the contract. The New York Fed study addresses the treatment question with within-firm comparisons, but its calibrated model must still account for why firms enter procurement in the first place. Selection is not a nuisance to be removed after the analysis; it is part of how the procurement system works.

The model separates the participant premium into two dimensions. In its calibration, procurement firms are 43 log points more productive and hold 83 log points more net worth before treatment. Those figures are model-generated interpretations rather than a directly observed decomposition. Their value is conceptual: productivity affects what a firm can deliver, while net worth affects what the firm can finance. The two may reinforce each other, but they are not interchangeable.

A productive firm with little net worth may have a profitable project in reach but be unable to prepare a proposal, hire workers, or buy inputs before payment.