significance for many estimated treatment effects.
1
Anderson (2008) claims that the program
does not work for boys, although he examines only a subset of its outcomes using arbitrarily
constructed indices of diverse outcomes, and he does not perform a cost-benefit analysis
overall or by gender.
2
The existing cost-benefit analyses of the program do little to assuage
these concerns, presenting estimates of rates of returns without standard errors, leaving
readers uncertain as to whether the estimates are statistically significantly different from
zero.
3
In response, Heckman, Moon, Pinto, Savelyev, and Yavitz (2010b) present the first
rigorous cost-benefit study of the Perry program that addresses four major challenges: (a) the
compromise inherent in the randomization protocol (see Heckman, Moon, Pinto, Savelyev,
and Yavitz, 2010a); (b) the lack of program data past age 40 and the need to extrapolate
out-of-sample to obtain earnings profiles past that age to estimate lifetime impacts of the
program; (c) missing data for participants before to age 40; and (d) the difficulty in assigning
reliable values to nonmarket outcomes, such as crime. The last point is especially relevant
for any analysis of the Perry program because crime reduction is touted as one of its major
benefits. This paper summarizes the main findings from our study. For more detailed
discussion of the results summarized here, see Heckman, Moon, Pinto, Savelyev, and Yavitz
(2010b).
Table 1 presents the range of estimates from our preferred methodology, defended in
Heckman, Moon, Pinto, Savelyev, and Yavitz (2010b). It supports separate rates of return
for benefits accruing to individuals versus those that accrue to society at large. Our estimate
of the overall social rate of return to the Perry program is in the range of 7% to 10%. We
report a range of estimates because of uncertainty about some components of benefits and
costs that cannot be quantified by standard errors alone. These estimates are above the
historical return to equity,
4
but generally below estimates reported in previous studies.
1
See Herrnstein and Murray (1994, pp. 404–405) and Hanushek and Lindseth (2009).
2
See Heckman, Moon, Pinto, Savelyev, and Yavitz (2010a), who use small sample permutation tests and
multiple hypotheses testing methods to establish that there are strong treatment effects for boys and girls,
although their life-cycle realizations differ across groups.
3
See Rolnick and Grunewald (2003) and Belfield, Nores, Barnett, and Schweinhart (2006).
4
The post-World War II stock market rate of return on equity is 5.8% (see DeLong and Magin, 2009).
3