Sometimes, you will hear a governor talk about the need to entice more entrepreneurs to his state. In my opinion, if you really want to bring entrepreneurs into your state, you should enact a school voucher program that enables entrepreneurs to compete with public schools on a level playing field.

Since I sold my business, I have said that I have no plans to become a serial entrepreneur. However, I am attracted by the notion of starting a school. Unfortunately, Montgomery County, Maryland is so heavily Democratic that even charter schools are viciously opposed by the local authorities.

As an economist, I tend to look at education from the standpoint of efficiency and equity. This essay will examine the case for vouchers in terms of efficiency, which means the quality of education relative to cost. My next essay will look at the issue of equity, which means the role of education in affecting the distribution of income and wealth.

Our youngest daughter currently is in sixth grade in the local public schools. This is her first year of "middle school." Recently, the school held its annual parent night, where we got to meet the teachers. The science teacher mentioned casually that over the course of a day she teaches 180 students.

I was surprised by the figure of 180 students per day. From an educational perspective, this seems like too many. How can a teacher assign and grade homework from 180 students?

However, what really struck me about the teaching load was its economic implications. It made me think in terms of the key operating ratios in Montgomery County Public Schools.

There are about six academic classes in a day in middle school. Therefore, the middle school science teacher handles 1/6th of the total teaching for each of her 180 students. So we can say that she accounts for 30 "student-years" worth of education.

The county spends about $8000 per year on each student. Multiplying 30 student-years by $8000 gives $240,000, which represents the annual "sales per teacher" for the County.

Meanwhile, the cost per teacher, including salary and benefits, cannot be more than $80,000 a year. In other words, the County's primary cost is $80,000, and they mark it up all the way to $240,000. What a business! A consulting firm would love to have those kind of profit margins.

For comparison, consider Sandy Spring Friends School, where our oldest daughter goes to high school. Tuition is $14,000 a year, which at first makes it appear more expensive. However, the number of students that a teacher sees in a day is much lower. I believe that 60 per day is a conservative estimate. Again, we take 1/6th of that to get "student-years," because there are about six academic periods per day. This gives us 10 student-years per teacher at Sandy Spring. Multiplying by tuition, "sales per teacher" amounts to something like $140,000 per year.

Teachers are paid less at Sandy Spring, so that the cost per teacher might be just $60,000, including salary and benefits. Nonetheless, it is not as good a business as the public school. The public school charges a $160,000 markup over its direct cost ($240,000 - $80,000), while Sandy Spring only has a profit margin per teacher of only $80,000 ($140,000 - $60,000).

Of course, what I am calling "markup" or profit margin actually is overhead. My point is that public schools have a lot of overhead.

The County will tell you that the overhead is necessary to handle "special needs" and diversity. In fact, there clearly are programs in the budget that reflect these issues.

For example, the western half of the County traditionally has been more affluent and less diverse than the eastern half, where I live. These disparities give rise to special needs.

The special need that the County identified was that the affluent white kids from the western half require enrichment programs, such as advanced mathematics and science. They also need to visit schools in our half of the county in order to make their parents feel less guilty.

To meet these special needs, the County set up "magnet" programs in some of the schools in our half of the county. The rich white kids who require advanced mathematics instruction are brought over in buses every day to their special math programs. At lunch, they get to see our "diverse" population.

My guess is that this is not the most efficient way to meet the special needs of the affluent white kids and their parents. Instead, the school system might have offered the special math programs in the local schools in the western half of the County, and sent the kids on occasional field trips to our half in order to see children of other backgrounds.

More generally, what if we had a voucher system in place in Montgomery County? What would happen to private schools, public schools, and teachers?

Let us assume that the public schools all would be taken over by private entrepreneurs. There might emerge a variety of school formats, but I want to speculate about the average school.

My guess is that the "profit margin" or "markup" would fall. It would fall for the former public schools, because of competitive pressure. It would fall for private schools, because they would tend to become larger, helping to create economies of scale. (Some classes at Sandy Spring have only a handful of students, which is educationally unsound as well as economically unsustainable.) Let us assume that the markup per teacher would settle at $60,000 per year.

The number of student-years per teacher probably would end up somewhere between the 10 that we see at Sandy Spring and the 30 that we see in public school. I would hope that it would end up closer to the Sandy Spring figure. Let us say that it turns out to be 15.

If the number of student-years per teacher ends up at 15, then we will need many more classroom teachers than currently are supplied by public schools. Some of these teachers would come from cutbacks in overhead, as administrators return to the classroom. However, many of the new teachers would have to come from the market. This would require higher salaries. Let us assume that the total cost per teacher would have to rise to $90,000 per year.

Adding the teacher salary of $90,000 to the markup of $60,000 gives $150,000 per year as the basic operating cost per teacher in our hypothetical average post-voucher school. If a teacher handles 15 students, then to get $150,000 we need to charge $10,000 a year. If students receive only an $8000 voucher, then on average families will have to pay an additional $2,000 a year on top of their vouchers. In other words, we cannot double the number of teachers per student simply by increasing efficiency. More money must be spent as well.

Opponents argue that vouchers cannot possibly work. As an economist, I find their arguments untenable. Markets work, and prices adjust. The bloated overhead of public schools (and, to a lesser extent, of private schools) can be cut. Teacher salaries can be increased. And parents can supplement public expenditures.

You will hear people scorn vouchers by saying that "there are not enough private schools." The point of this essay is that with plausible adjustments the market could make every school a private school.