21.07.2013 Views

The State of Minority- and Women- Owned ... - Cleveland.com

The State of Minority- and Women- Owned ... - Cleveland.com

The State of Minority- and Women- Owned ... - Cleveland.com

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

(2) Ii = β0 + β1CWi + β2Xi + β3Ri + β4LCi + εi,<br />

Statistical Disparities in Capital Markets<br />

where I represents the interest rate charged on the loan, LC represents characteristics <strong>of</strong> the loan<br />

(see the notes to Table 6.8 for a full list <strong>of</strong> the variables included in this set), εi is a term<br />

capturing r<strong>and</strong>om factors, <strong>and</strong> all other notations are the same as in equation (1).<br />

An important consideration is whether the interest rate may be treated as exogenous, as our<br />

reduced form model assumes. In the context <strong>of</strong> small business loans, in which it is possible that<br />

the loan terms may be negotiated in the determination process, this assumption may not be valid.<br />

As such, a model that simultaneously estimates the interest rate <strong>and</strong> the loan decision might be<br />

appropriate, except that the interest rate that would be charged to firms whose loans were denied<br />

is not available in our data. Alternatively, one could estimate an interest rate model alone for<br />

those firms whose loan was approved, adjusting for the potential bias brought about by sample<br />

selection. To properly identify such a model, however, a variable is required that is linked to the<br />

loan denial decision, but unrelated to the level <strong>of</strong> interest charged on approved loans; no such<br />

variable exists in the data.<br />

Nevertheless, one would expect these considerations to impose a downward bias on the<br />

estimated differential in interest rates charged on loans to African American-owned firms. Those<br />

firms whose loans were rejected would have been charged higher interest rates than those<br />

approved. Since African American-owned businesses were considerably more likely to be<br />

rejected holding constant differences in creditworthiness, one would expect any differential in<br />

interest rate to be even greater if those firms were included in the sample. We overlook this<br />

implication in the results reported below, but its impact should be kept in mind.<br />

<strong>The</strong> results obtained from estimating equation (2) are reported in Row 1 <strong>of</strong> Table 6.13, which<br />

includes the <strong>com</strong>plete set <strong>of</strong> control variables <strong>com</strong>parable to those in Column 5 <strong>of</strong> Table 6.8.<br />

Estimates indicated that African American-owned firms pay rates <strong>of</strong> interest that are roughly one<br />

percent (100 basis points) higher than similarly situated nonminority-owned firms. Row 2 shows<br />

that even African American-owned firms with good credit histories are charged higher interest<br />

rates relative to nonminority-owned firms. 285 <strong>The</strong>re is some evidence <strong>of</strong> this phenomenon<br />

affecting Hispanic-owned firms as well.<br />

<strong>The</strong> remainder <strong>of</strong> the table presents similar specification checks to those reported in Table 6.10.<br />

Recall that most <strong>of</strong> these models identify firms for which the firm’s own history is likely to be a<br />

more important contributor to its creditworthiness. <strong>The</strong> specifications by sales market are<br />

designed to distinguish the impact <strong>of</strong> central city location. Unfortunately, sample sizes are<br />

smaller in these specifications <strong>and</strong> reduce the power <strong>of</strong> the analysis. Nevertheless, we still find<br />

that regardless <strong>of</strong> organization type <strong>and</strong> firm age, African American-owned firms face<br />

statistically significantly higher interest rates. Overall, the evidence presented indicates that<br />

African Americans, <strong>and</strong> to a lesser extent Hispanics <strong>and</strong> Asians/Pacific Isl<strong>and</strong>ers, face<br />

disadvantages in the market for small business credit that do not appear to be attributable to<br />

differences in geography or creditworthiness.<br />

285 Estimates from firms that have had past credit problems are not presented since the higher likelihood <strong>of</strong> their<br />

being denied credit restricts the size <strong>of</strong> the sample <strong>and</strong> limits the ability to provide a powerful test <strong>of</strong> the interest<br />

rates charged if they are approved.<br />

NERA Economic Consulting 207

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!