Over the years, I’ve written frequently about government efficiency and inefficiency. I even wrote a book that, among other things, measured types of efficiency across 38,000 local governments in the U.S.

We now live in a time when lots of folks complain about government efficiency. Some of them have experience in government and some have built large private sector firms, while others haven’t been in charge of anything important—often for good reason.

It is helpful to think about what might cause government efficiency or inefficiency, and whether it is really the problem that we think it is. I’ll begin with a simple observation that there a number of things governments shouldn’t be doing. The real standard for this is whether the private sector does it somewhere.

By that standard, governments shouldn’t run airports or sports arenas. They shouldn’t own or operate hotels, restaurants or hospitals. Governments shouldn’t buy land for private development or operate loan funds for businesses. There may be rare exceptions to all these rules, and misuse of public funds in all these examples is a real concern. However, that isn’t the point of this column.

Here in the U.S., we employ a federalist form of government. That means different levels of government concern themselves with different problems. The federal government should be minimally involved with local schools, perhaps collecting data or paying for research that everyone can learn from. City councils should be silent on foreign policy.

There are three clear causes of government inefficiency.

The first is scale. In my 2012 book (with colleague Dagney Faulk), we found considerable inefficiencies due to government operations being stuck at an inefficient size. Economists since Adam Smith have noted that there is an optimal size for most production processes. It changes over time, and with technology, but there is always a range of maximum efficiency.

The best example of this is in fast-food restaurants. No matter where you go, in any city, town or country, fast-food restaurants are always built and operated in similar sizes. This isn’t an accident. Businesses choose the optimal size of operations to maximize profits and make investment and hiring decisions accordingly. That’s the benchmark for efficiency.

Public sector entities don’t really have that freedom.

Across most of the country, there are lots of tiny school districts that are inefficiently small and big ones that are inefficiently large. The same is true with police, fire, administration, parks, libraries and other functions. The people making budget decisions in these places largely have no meaningful say in the scale of their operation. So, that inefficiency is baked into the facts surrounding their jobs.

What is most interesting is that the small-scale inefficiencies of this type are mostly clustered in rural places, and rural voters seem perfectly happy to let them continue that way.

Inefficiencies in larger-scale operations—like big-city schools in New York or Chicago—appear to be the result of legislative action. So, inefficiencies in these places aren’t about scale, but are about city councils or state legislatures making rules that create inefficiencies to protect special interests. I don’t like them, but voters seem content to allow them to persist.

The second source of inefficiency in government operations is their complexity in comparison to the private sector. In most U.S. cities, the largest fleet of vehicles is operated by the local school. The largest restaurant service is the high school or elementary, and the most highly attended paid venue is the local basketball game. That same school corporation has the largest local maintenance contracts, the largest computer array, the largest library and the largest HVAC system in the county.

That same school system has the most challenging physical security requirements in the county and must make weather-related cancellation decisions about 50 days a year—before 5:30 a.m. And these decisions are calibrated by complex concerns beyond safety. For example, a large share of students get their only breakfast at school.

I have enormous respect for the private sector and the great wealth it creates. We all should. But there are precisely zero examples of private sector firms doing anything like this today.

The third source of inefficiency in government is the absence of a profit to measure success. Government measures of success, from the battlefield to the classroom, the public pool and the library, are fuzzy and imprecise. Not so for a bank or restaurant.

We may not know how good our schools are, how important USAID spending is, or even if we have successfully prosecuted a war, for decades. Not so for a manufacturing or logistics firm, or a fast-food restaurant.

Indeed, how do you measure how good a rifle battalion of 800 soldiers is? How useful is foreign aid in country that might otherwise become a hotbed of terrorists? Or how good the parks department might be, or the library or police department? It is easy to generate metrics for each of these. We do so, everywhere, to a stunning extreme. That is what 100 percent of school testing is about.

Improving government efficiency is an important goal that requires serious thinking from serious people. When you hear an elected official or media personality rail on about government inefficiency, but they cannot explain how and where it is systemically occurring, and how it can be remedied without cutting tax, you might conclude they aren’t serious.