Measuring SBLF’s Impact Using Advanced Statistical Tools

FI Consulting

Measuring SBLF’s Impact Using Advanced Statistical Tools

FI Consulting helped the US Department of the Treasury quantify the policy impact of the Small Business Lending Fund (SBLF) program and demonstrate their role in increasing small business lending.

Client: US Department of the Treasury

Challenge: The Small Business Jobs Act of 2010 directs Treasury to report on how SBLF participating lenders have used the funds they received. To measure SBLF’s impact, each quarter the program reports changes in small business lending relative to baseline levels before it took effect. These reports compare SBLF participant lending activity to those of peer institutions and comparable non-participating lenders.

While the selection of peer and comparable institutions considers variables such as geography, size, and financial condition, any such selection may be subject to the effect of unobserved factors that could simultaneously influence an institution’s decision to participate in SBLF and increase business lending. Selecting properly balanced peer and comparable institutions while controlling for selection bias is critical to properly measuring project impact.

*FI Solution: *By applying a social sciences technique commonly used in healthcare drug trials, FI Consulting helped Treasury create more rigorously grounded estimates of SBLF’s effect on small business lending. The propensity score matching analysis allowed SBLF to create a statistically similar control group that better isolated the “treatment effect” of a bank participating in the program. Taking an innovative approach in the application of this methodology to measure the impact on lending growth, FI and Treasury worked together to apply two approaches to evaluate SBLF’s impact. The first approach matched SBLF banks to a control group based on their propensity to participate in the SBLF program and the second adjusted for banks’ additional propensity to raise capital from non-government sources. To isolate the effect of the program from other external factors, FI collected a robust dataset of over 70 covariates for the entire population of small- to mid-sized US community banks. Assembling two cohorts of SBLF and non-SBLF lenders in this manner mimics the effect of randomization, permitting a clearer view of the effect of SBLF participation as a key driver of increasing the availability of credit to small businesses.

*FI Impact: *By applying these analytical techniques, FI Consulting helped Treasury reaffirm that SBLF has been effective in increasing lending to small businesses in the wake of the financial crisis. The FI Consulting team’s longstanding partnership with Treasury and its core strengths in data, analytics, modeling, and technology helped the federal government better understand and measure the impact of its effort to help small businesses obtain the funding they need to grow and strengthen the US economy.