In chapter 12 of Investing in Kids, I also consider the economic development benefits of improvements in test scores in secondary school. These economic development benefits are the increase in a state’s per capita earnings. These increases in per capita earnings will occur as some of these better secondary students stay in the state, and help the state attract and grow more and better jobs.

To analyze state economic development benefits of improving secondary test scores, I use a similar methodology to what I used to estimate the economic development benefits of early childhood programs, or of elementary test scores in a previous post. I use data on how secondary school test scores are related to adult earnings. I also use estimates of how many high school students will stay in the same state during their working years. I use estimates of how improved labor force quality will affect job and earnings growth.

The secondary score test score improvement that is considered is an increase of 0.1 in “effect size units”.  The “effect size” educational jargon measures test score improvements as a fraction of the standard deviation of test scores across a regular cross-section of students. For secondary schools, test scores vary quite a bit across students. An effect size of 0.1 corresponds to what the average student in high school normally learns in about 5 months, or in about one-half of a school year.

The previous post on elementary school test score improvements also considered an improvement in test scores of 0.1 in effect size units. However, test scores do not vary as much across students in early elementary school. Therefore, this improvement at an elementary level corresponds to only 1 month or so of extra achievement.

An effect size increase of 0.1 at the secondary level is probably more costly to achieve than the same effect size increase at the elementary level.  It is probably harder to increase test scores by 5 months at the secondary level than by 1 month at the early elementary school level.

The secondary test score increase I consider is for one student. I estimate that the state economic development benefits of a five month increase in test scores at the secondary level for one student will be $7,050. This is the increase in the state’s earnings per capita that is associated with this increase in test scores for one student.

A benefit of $7,050 is sufficient to justify a costly investment. We could imagine a variety of high school reforms that might improve test scores by the requisite amount at a cost per student of less than $7,050.

However, the ratio of economic development benefits to costs for many secondary school reforms will be less than the ratio for many elementary school reforms.  As mentioned in a previous post, the benefits of a 0.1 effect size improvement for one student at the elementary level are slightly higher, at $8,312. And this elementary school test school improvement will often be less difficult to achieve than a similar effect size improvement at the secondary level.

Why are earlier K-12 investments likely to have higher returns?  The biggest reason is that relatively modest test score improvements at the elementary school level are more strongly associated with adult earnings gains than is true for secondary school test score gains. Test score gains at the elementary level may be more strongly associated with improvements in soft skills. Furthermore, skills may be more malleable for younger children.

As pointed out previously, we might do still better by investing even earlier, in high quality preschool, parenting, and child care programs. This does not mean we should not do later investments. These later investments are needed to take full advantage of the earlier investments. But we should put a high priority on the earlier investments given their higher returns.