Overview

The Census evidence separates three margins after state affirmative-action bans in five treated states: existing MWBEs shrink relative to non-MWBEs, fewer new MWBE contractors enter, but observed incumbents are not detectably more likely to exit. A partial-equilibrium model can reproduce the employment and exit pattern with an outcome-equivalent 8.6% decline in median long-run productivity and 10% lower scrap value, yet these are calibrated interpretations, not direct measures of technology or sale prices.

Learning outcomes

Fewer new MWBE contractors entered after bans, while observed incumbent MWBEs were not detectably more likely to exit.
Employment and operating scale can fall without closure, so exit rates alone understate lost capacity or opportunity.
The model's 8.6% productivity and 10% scrap-value declines are outcome-equivalent calibrations, not direct measurements.

Keywords

ProcurementMWBEEntryExitEmploymentScrap value

Shrinking Without Closing

Government procurement can change the lifecycle of a business in more than one way. This chapter focuses on the distinction between an incumbent contractor becoming smaller, a firm exiting, and a potential contractor never entering after access changes. Keeping those margins separate lets the reader interpret the Census evidence correctly and understand why its calibrated measures of productivity and firm value are outcome-equivalent translations rather than direct observations.

Entry falls, but exit does not rise

The Census study follows two related but different populations. For exit, it starts with firms identified in the 2002 or 2007 Survey of Business Owners and tracks their linked Longitudinal Business Database records. For entry, it uses the 2017 Annual Business Survey, whose respondents could not have exited before the survey year. This split is necessary because a snapshot of firms has different information about what happened before and after the snapshot. It also means that entry and exit estimates answer different questions and should not be read as two measurements of one complete population flow.

The exit result is notably restrained. Existing MWBEs are not statistically more likely to leave than comparable non-MWBE contractors after a state bans affirmative action. All exit coefficients are indistinguishable from zero at the 95% confidence level; at lower confidence levels, some estimates point toward MWBEs being less likely to exit. The strongest safe reading is not that the bans had no survival effect on any firm. It is that the study does not detect a relative increase in exit for the observed incumbent population with the available data and uncertainty.

That finding sits alongside the employment decline. A firm can reduce its workforce, accept fewer projects, postpone equipment purchases, or serve a smaller customer base while continuing to operate. Exit is a threshold decision: it occurs only when the value of continuing falls below the value of stopping, selling, or otherwise disposing of the business. A loss of procurement access can reduce the scale and expected future value of an incumbent without pushing that value below the exit threshold. Stable exit is therefore compatible with economic damage or reduced opportunity.

The entry result points in a different direction. After a ban, relatively fewer new MWBE contractors enter compared with non-MWBE contractors. The fifth-year estimate implies roughly one fewer MWBE entrant per state-by-sector cell. Using the study's 20 two-digit sectors and an adjustment for overlap between minority- and women-owned categories, the author calculates that a treated state loses approximately 35 new MWBE contractors per year five years after the ban. That figure is a back-of-the-envelope extrapolation, not a directly counted statewide total, but it conveys the scale implied by the cell-level estimate.

Entry is forward-looking in a way exit is not. An incumbent has an existing organization, customer relationships, knowledge, and sunk investments. A potential entrant compares the cost of starting with the expected value of becoming an operating contractor. If a policy removes a route to government work, the person or firm considering entry may never appear in the data. The resulting decline in new contractors can therefore be economically important even when the visible incumbent population remains in business.

Published by DeepMox

FAQ

Why can a firm shrink without exiting?

A firm can keep operating with fewer workers, projects, or investments when lost procurement access lowers its continuation value but not below the exit threshold; closing or selling may also become less attractive.

How do entry and exit estimates in the Census study differ?

Entry uses 2017 Annual Business Survey respondents to measure new contractors after bans, while exit tracks firms identified in the 2002 or 2007 Survey of Business Owners. They describe different populations, not one complete flow.

What does the 8.6% productivity estimate mean?

It is an outcome-equivalent decline in median long-run MWBE productivity under calibrated model parameters that reproduce observed employment and exit patterns, not evidence that every firm's technology fell by 8.6%.

Previous9 minState Government Procurement Preference Bans and MWBEs

Next11 minGovernment Procurement: Small Suppliers and Public Output

Government Procurement: Shrinking Without Closing

Shrinking Without Closing

Entry falls, but exit does not rise

FAQ