CONVICT LABOR FOR ROAD WORK. 17 
in other States the work done under the State-use system, the only 
other system which does not conflict directly with the interest of free 
laborers, is found to be even less efficiently performed than is the road 
work. Furthermore, except in those sections where the wages of free 
laborers are exceptionally low or the efficiency of such labor excep- 
tionally high, there seems to be no good reason why road work can 
not be accomplished by convict labor at considerably less expense 
than by free labor. 
EFFICIENCY AND ECONOMY OF CONVICT LABOR. 
The relative efficiency of convicts and free men as road laborers is 
a phase of the convict problem of particular interest. Unfortunately, 
it is also a phase upon which it is practically impossible to develop 
precise information. Manifestly, an entirely fair comparison can be 
made only where both classes are employed in like localities under 
exactly similar conditions. This is rarely possible, because convicts 
and free men are seldom employed together, even on different sec- 
tions of the same road where conditions might be assumed to be 
roughly identical, but by making proper allowance for differing con- 
ditions it is sometimes possible to form reasonably accurate estimates 
of the comparative value of the two classes of labor. Estimates of 
this sort are not wanting, but in their bearing on the general question 
of the efficiency of convict labor they serve to confuse rather than to 
illuminate, for they rate the relative efficiency of the convict at from 
50 to 150 per cent of that of free labor. 
By assembling a number of such estimates from different localities 
and under different conditions it is possible to arrive at a composite 
figure which will represent the average relative efficiency of convict 
labor throughout the localities represented. An estimate of this sort 
was made by the United States Bureau of Labor and published in the 
Twentieth Annual Report of the Commissioner of Labor in 1905. 
The data for that estimate were secured by agents of the Bureau of 
Labor from prison officials, foremen, contractors, lessees, and from 
employers of free labor in the localities in which convicts were em- 
ployed. As a result of this survey, it was found from a total number 
of 111 estimates in regard to highway construction in the States of 
California, Connecticut, Florida, Georgia, Kansas, Michigan, Minne- 
sota, Missouri, New Jersey, New Mexico, New York, North Carolina, 
Oregon, Pennsylvania, South Carolina, Texas, Virginia, Washington, 
and the District of Columbia, that the labor of 3,522 convicts was 
equivalent to that of 3,481 free laborers of average skill, working the 
same number of hours per day. But that such general estimates 
are of little value in the consideration of particular cases is well 
illustrated by the fact that if the estimates for the States of Cali- 
