Project Overview
This research project examines the causal effect of Colorado's HB17-1214, an employee ownership support program enacted in 2017, on firm-level ESOP participation and asset accumulation.
Access the Full Report
Key Findings
Our analysis using Department of Labor Form 5500 filings (2012-2025) with a difference-in-differences design comparing Colorado to Utah reveals:
- Limited policy effects: Small and mostly statistically insignificant increases in ESOP participation and assets
- Marginal significance: Only total ESOP assets showed marginal significance at the 10% level (~$2.9 million increase, representing ~9% relative to baseline)
- Participation outcomes: ESOP firm indicator increased 0.6 percentage points (insignificant); total participants and active participants showed ~9% point estimates but lacked statistical precision
- Synthetic control analysis: Suggested larger effects, but poor pre-treatment fit undermines confidence in these estimates
Our findings suggest education-based and financing-focused employee ownership policies may require longer time horizons or more direct financial incentives to generate substantial ESOP adoption among small and medium-sized businesses.
Policy Context
Colorado's HB17-1214 (signed by Governor Hickenlooper in 2017) focused on:
- Educational support through contracted nonprofit organizations specializing in employee ownership
- Loan program capitalized with $200,000 annual state appropriations (2017-2022) to finance ESOP transitions
- Eligibility: Small businesses (< $5 million revenue), at least 2 years old with ≥3 employees
Reported outcomes: 20-25 businesses per year transitioned to employee ownership with program support.
Methodology
Difference-in-Differences (DiD)
We compare Colorado firms (treatment, post-2017) to Utah firms (control) and Colorado firms (pre-2017) before and after the 2017 policy implementation. This removes time-invariant differences between states and common macroeconomic shocks.
Identification Assumptions
- Parallel Trends: Event study analysis confirms Colorado and Utah exhibited parallel trends during pre-treatment period (2012-2016)
- Compliance: HB17-1214 provisions exclusively available in Colorado; Utah enacted no comparable policies
- SUTVA (No Spillovers): Program benefits require Colorado residency; ESOP transitions unlikely influenced by neighboring state policies
- Exclusion Restriction: Policy narrowly targeted education and ESOP-specific financing with modest funding; unlikely to generate substantial spillovers
Robustness Check
Synthetic control methodology constructing counterfactual "synthetic Colorado" from weighted combinations of other U.S. states (Iowa and Missouri excluded due to own ESOP legislation).
Important Limitations
- Inference constraints: Treatment assigned at state level while outcomes measured at firm level; with only two states, state-level clustering is infeasible, potentially overstating precision of firm-level estimates
- Limited statistical power: Single treatment state and control state provides minimal variation to distinguish treatment effects from Colorado-specific shocks
- Confounding variables: Pension plan participation depends on firm profitability, employee turnover, competing benefits, and management preferences—state-specific time-varying factors cannot be controlled
- Data limitations: Form 5500 lacks detailed firm characteristics; cannot examine heterogeneous effects by firm size or ESOP structure
- Missing coverage: Analysis may miss smaller ESOPs below Form 5500 reporting thresholds (typically plans with 100+ participants or more complex structures)