We are able to observe the participation rate of households in the MSFH program by zip code.
Neighbor Effect Hypothesis: The likelihood of self-insurance is positively related to the self--insurance behavior of nearby households (as proxied by area participation rates in the MSFH program).
Table 2 provides the summary statistics for the individuals in our sample and values for individuals who reported taking self-insurance action and those who did not for the sample of homes eligible for the MSFH Grant program.
As described above, our sample is based on homeowners who participated in the MSFH program.
Given that the MSFH grant program only provides subsidized grants for homes under $300,000 in value, we first run the models based on a subsample of homes valued at $300,000 or less.
Among the control variables, the positive coefficient for WBDR is not surprising because (1) homes located in the wind-borne debris region are at higher risk for hurricane damage, and (2) an MSFH program requirement for receiving a grant was that a home was located in the WBDR.
Based on the first-stage probit model, a representative household in our study that took part in the MSFH program is predicted to have a 6.2% probability of making mitigation improvements, holding the independent variables at their mean values.
For robustness, we also examine results for the full sample of homes (i.e., including those homes insured for amounts greater than $300,000 and therefore not eligible for the MSFH program).
Results suggest, consistent with the model in Kelly and Kleffner (2003) showing that individuals will engage in more mitigation when government subsidizes the cost, initiatives such as the MSFH grant program can increase levels of mitigation.
Florida Division of Finance Services, 2007, IRS Grants CFO Sink's Request That MSFH Grants Not Be Subject to Federal Income Tax.
For more detailed information on the My Safe Florida Home program, see the MSFH Annual Report (2008) and CFO Bulletin (2007).