Economist Writing Every Day

Who knows what state capacity was 150 years ago? After all, DMV jokes are only a little out of date. There is a lot of richness and specificity to state capacity. That’s why we can’t look at an identical law in two different places or times and assume that their enforcement and evasion are the same.

Interested readers can see my previous post for a figure that illustrates the timing of compulsory education legislation across US states. The effects on literacy were a bit ambiguous. The explanation might be that effective enforcement by the various states might have differed (substantially). The figure below illustrates the average rates of attendance by age and census year.

Just as an increasing number of states began to enact compulsory school attendance, we can see that school attendance rates rose over time. But we can’t tell from the figure whether attendance laws caused or were merely coincident with increasing attendance.

One hint is to group the people by whether their state had compulsory attendance laws on the books. See the figure below.

No state had compulsory schooling in 1850 and all states in the sample had compulsory schooling by 1920. What stands out most to me is that, decade by decade, states with compulsory attendance laws had consistently higher attendance (no surprise). Another pattern is that the difference shrinks over time. In 1870 there was up to a 20 percentage point difference. By 1910 the difference was more like 6 percentage points.

In fact, attendance over time is underestimated in the latter decades because the details of the census question changed. For the censuses in 1850-1900, school attendance in the past 12 months was counted as a ‘yes, attended’. However, attendance in 1910 was counted in the past 7.5 months and then in the past 4 months in 1920. But, it is impressive that the attendance rates rose by so much in 1910 and 1920 despite the narrower definition of attendance.

There is still a problem. States which had the highest attendance rates prior to legislation would also have the lowest enforcement costs – and therefore, legislators would be willing to enact compulsory attendance sooner. Such a hypothesis is falsifiable with some more in-depth analysis… another time. For the moment, I’m willing to believe that, even 150 years ago, states were able to force kids into schools (as far as the parents who answered the survey knew, anyway).

PS – My IPUMs sample includes only people designated with a race of ‘white’ in order to avoid other effects.