|
Coefficient units are additional annual
crime incidents per 100,000 population.
t-statistics are in parentheses.
We used robust standard errors for larceny and
burglary, which the Breush-Pagan test indicated
had
heteroscedasticity. .
For aggravated assault, the coefficient
estimates for both years of leads, the year of
opening,
and the first two years after the casino opening
are not statistically different from zero.
However,
estimates for the third and subsequent year
after opening are significantly above zero, and
the
trend rises. By the third and subsequent year
casinos are a statistically significant
contributor
to increased assault rates. The estimated high
occurs in the fifth year after opening, when the
aggravated assault rate is 100 assaults higher
per year.28
28The estimated pattern of crime increase is
unlike the typical pattern of visitor increases
after casino opening.
23
Notes: Coefficient units are additional annual
crime incidents per 100,000 population.
t-statistics are in parentheses.
We used robust standard errors for larceny and
burglary, which the Breush-Pagan test indicated
had
heteroscedasticity. .
Figure 5.2 for rape shows a pattern similar to
aggravated assault. Coefficient estimates are
Grinols and Omorov (1996) showed the number of
visitors to Illinois casinos typically rises
immediately after
opening and reaches equilibrium levels after six
months or fewer.
In addition to the regressions reported, we also
ran regressions that included as many as 4 leads
and up to 7
years of lags of the casino opening variable.
The results tended to be extensions of the
patterns of the reported
regressions. That is, with few exceptions leads
continued the pattern of being statistically
indistinguishable from
zero and later lags showed comparable or greater
estimated effects. In the case of murder, the
6th and 7th lags
continued the pattern of being statistically
indistinguishable from zero.
Back to casino
news
|
