Does Discretion Enable Active Representation? Evidence from the 2022 Expansion of WOSB-Eligible Industries

Overview

A foundational claim in the study of representative bureaucracy holds that discretion is what turns passive representation into active representation. The claim is widely invoked, yet rarely tested causally. This paper offers such a test. I use the 2022 expansion of WOSB-eligible NAICS codes as an exogenous increase in contracting officers’ discretion and, drawing on federal procurement and workforce data from 2015 to 2024, ask whether more discretion alone changes behavior. It does not. Discretion increases the use of WOSB set-aside tools only where a contracting office already has a higher share of female managers. Active representation, in other words, emerges where identity-based willingness meets institutional opportunity—and neither component does much on its own.

This pattern reveals an activation mechanism:

Active Representation = Willingness (passive representation) × Opportunity (discretion)

Research Design

• Policy Shock (treatment):
  2022 Small Business Administration expansion increased the number of NAICS codes eligible for WOSB set-asides, exogenously expanding bureaucratic discretion.

• Data:
  • USAspending / FPDS procurement data (2015–2024)
  • OPM STATUS workforce records to identify female managers in the 1102 contracting series
• Empirical Strategy:
  • Difference-in-Differences (DiD) comparing newly eligible NAICS codes vs. never eligible codes
  • Triple Difference (DDD) interaction with pre-2022 female manager share
  • Outcome Variables:
    • Bureaucratic intention: WOSB set-aside usage
    • Market outcomes: Final awards to WOSB firms

Key Findings

1. Discretion alone has no average effect.

The policy expansion does not increase WOSB set-aside usage on average (ATT ≈ 0).

2. Discretion activates representation only when female managerial presence is high.

Triple-DiD shows a positive and significant interaction:

• Agencies with more female contracting managers used WOSB set-aside tools substantially more after discretion expanded.
• Agencies with few female managers showed no behavioral change.

3. Activation occurs for bureaucratic intention, not market outcomes.

• Set-aside behavior shows activation.
• Final WOSB award shares do not exhibit the same moderated effect.

This distinction locates representational behavior where it actually occurs: inside bureaucrats’ decisions, not necessarily in market results shaped by competition.

Theoretical Contribution

This study offers causal evidence that discretion is not a simple main effect; it is the mechanism that activates latent representational motivations. Passive representation (female managers) matters only when the organization grants bureaucrats meaningful choice. Without willingness, opportunity has no effect; without opportunity, willingness cannot be expressed.

The paper confirms active representation as the product of two necessary components:

Identity-based Willingness × Institutional Opportunity