In many parts of the country, the average cost of professional child care exceeds $20,000 per year — a fact which is motivating calls for new subsidies to reduce the cost of child care for low income families. And yet the high cost of child care has no easy fix. Before pouring money into existing child care providers, subsequently driving cost growth that would simply make the situation worse, we need to get the facts straight about the scope and nature of the problem.

Fact 1: Child Care Subsidies Will Worsen Cost-disease

Like many health care and educational services, professional child care services are relatively low skill and labor intensive. This makes child care subject to the cost-disease — the phenomena whereby productivity growth in one sector of the economy causes wages, and thus labor costs, to rise in similarly skilled sectors in order to attract and maintain employees. There are many reasons why the cost of child care has increased, but cost-disease is by far the most important. And it’s not about to go away.

The cost-disease poses an inherent problem for efforts to directly subsidize child care. To see why, consider that in a simple economic model, the inability to realize productivity improvements is represented by a supply curve that is inelastic in the long run. That means any increase in demand (which is what a subsidy represents) will eventually be translated into permanently higher prices, canceling out improvements in affordability. This process has been rigorously confirmed in the context of college tuition inflation, for example.

In normal goods markets, the long-run supply curve is typically elastic: an increase in the demand for widgets may raise the price of widgets in the short run. However, in the long run firms respond to the higher price by developing more productive techniques for producing widgets and compete the price back down to a long-run equilibrium. In contrast, in a cost-diseased sector like child care, there are natural (and often even legal) limits to how productive a child care worker can become, such as caps on the number of children a single worker can oversee at a time. Supply is also made less responsive through barriers to entry, from restrictive licensing requirements to quotas on low skill immigration.

Supply may nonetheless increase, but only by pulling in workers from more productive sectors through artificially higher prices and wages. You could call this approach “leaning into cost-disease,” and it presents the worst of two worlds. Not only do child care costs continue to escalate, but they come at the expense of the broader economy’s productivity.

Fact 2: Child Care Subsidies Will Slow Innovation

The effects of cost-disease are not limited to making low-skill, labor intensive services more expensive. Cost-disease also affects the relative costs of different institutional arrangements, putting pressure on old economic models to adapt or go by the wayside in ways that are often hard to foresee. The decline in community schools is the classic example. As teacher salaries grew commensurate with the wage growth in more productive sectors, the economics of small, local schools became untenable. Over time, cost-disease in education has caused thousands of community schools to consolidate into larger, regional schools that take advantage of economies of scale.

These sorts of transitions are always painful, but particularly so when they’re put off through subsidies and various kinds of supports that slow down cost restructuring. A major problem with subsidies to child care providers, whether done directly or indirectly through vouchers, is the risk of entrenching a particularly high cost model of child care delivery. This is why it is virtually always preferable to simply provide low income parents with cash, giving them the choice over different care arrangements based on market prices.

It is inherently difficult to predict the ways cost-disease will change the economy. In the case of health care and education, there is increasing pressure in the direction of telemedicine and online or hybridized college courses. It is probable that, in the near future, more and more jobs will be digitized, allowing telecommuting of various sorts. In a world of ubiquitous home offices, do external day care centers even make sense? The point is that no one can say.

Fact 3: External Care Arrangements Are Already the Exception

The question remains: How do lower income households afford the high cost of child care? The simple answer is that they don’t. As it stands, over 40% of children under the age of five are in regular care arrangements involving relatives, while less than 25% of children under five are in a formal, paid child care arrangement.

A report by the Center for American Progress last year called attention to so-called “child care deserts,” based on an analysis of the capacity of child care centers by zip code. The analysis suffers from some serious flaws, which we will be going in depth on in a forthcoming report. But as a simple conceptual point, the “desert” metaphor doesn’t work, when over 75% of preschoolers make use of alternative care arrangements (there are no alternatives to drinking water).

Indeed, the use of organized child care facilities has varied erratically over the years, even as maternal employment has steadily increased. While the most recent trend has seen an increased reliance on fathers and organized daycare facilities as primary care providers, a 2013 U.S. Census report notes that “the lack of a consistent trend since 1985 in the use of specific child care arrangements for preschoolers makes it difficult to foresee which arrangements will grow or wane in popularity in the future.”

Parents, on the whole, state a strong preference for relative or parental care over day care centers, which does more to explain the erratic uptake of external care arrangements than cost alone. And importantly, it’s a preference that is strongest in low income households. According to the U.S. Census,

low-income parents (those making no more than $25,000 annually) have more pronounced concern about day care centers than high-income parents (those making more than $75,000 annually). Seventy-two percent of low-income parents express great concern about the possibility of neglect or lack of supervision of children, compared with 51% of high-income parents. Seventy-six percent of low-income parents are more fearful that children may suffer physical or sexual abuse in day care, compared with 49% of high-income parents.

Results like this should give pause to would-be reformers. It suggests that the drive to push young children into child care facilities reflects a bias in elite attitudes, rather the preferences of the parents themselves. As I’ve written before, policies like a child allowance or an enhanced per-child tax credit would help low income families afford the costs of child care, in whatever form it takes.

Social engineering can take many seemingly benign forms. But when the objects of engineering are the choices families make in how they raise their child, it seems that a liberal dose of precaution is in order.