Some Thoughts About Services

Are Physical Goods Really Valued More Accurately Than Intangible Services?

It is sometimes argued that services like those provided by teachers, public administrators, and some technologies are “valued at cost” in economic statistics. The claim typically goes something like this: when we measure the value of a tangible good, we use its market price, which reflects both supply and demand. However, for many services, we often rely on the cost of labor (e.g., a government salary) because the service’s output is not directly priced.

From this, some infer that measures of economic well-being like Gross Domestic Product (GDP) capture the value of physical goods more accurately than the value of services. We track what we pay for the service, but that might only loosely reflect the societal value of the service’s output.

In what follows, I will argue that while these critics are correct to distinguish between the accounting cost of services (what is paid) and economic value (what society would give up to partake in the activity), this gap is not unique to services. Most market prices do not perfectly mirror social value. Furthermore, in the case of services, there is typically no strict need to separate the price of the “output” from the price of providing the service itself. The labor itself is typically the end product.

Economic Value vs. Accounting Cost

Let’s begin with what the critics get right. From an economic perspective, value depends on opportunity cost and the analysis of foregone alternatives. If you hire a teacher to give daily lessons, the economic value of that teacher’s labor depends on what society forgoes by employing them.

1. Opportunity Costs for the Employer (School or Government):
   The funds used to hire a teacher could alternatively purchase classroom supplies, fund field trips, hire additional administrators, or provide raises to existing staff. The wage offered to the teacher reflects, in part, that the institution believes the teacher’s services are more valuable than these other options.

2. Opportunity Costs for the Teacher:
   From the teacher’s perspective, other viable choices include teaching at a different school, working in a different field, or not working at all. The pay must therefore compensate the teacher enough to make this job at least as attractive as realistic alternatives.

In this way, a teacher’s wage bears a meaningful connection to the teacher’s opportunity cost and that of the employer. Of course, this connection is imperfect, since market failures and externalities can drive a wedge between the wage (accounting cost) and true social value. In our example, perhaps there are positive externalities to society from having a more educated population, which neither the teacher nor the school take into account. But such mismatches are not unique to services.

The Issue of Fees

One might argue that services are “valued at cost” because their outputs, such as student learning, are not separately priced. However, many services are directly priced. You might pay an hourly fee for guitar lessons, legal advice, or financial services. Those fees, like prices for physical goods, are influenced by competition, licensing, and other market and nonmarket forces.

Payments for goods and payments for services are both expenditures. Neither is guaranteed to reflect social opportunity cost accurately. If the government heavily subsidizes corn production, for example, the grocery store price probably does not faithfully capture the social value of that corn. Any number of taxes, subsidies, or externalities can distort the alignment of accounting cost and economic cost.

Ultimately, divergence between a market price and a theoretically optimal “shadow price” is near universal. The phenomenon applies to goods and services alike. Moreover, classifying a service payment as a “fee” versus a “salary” is largely a matter of semantics. Part-time or substitute teachers, for instance, might be paid on a per-class basis (functionally similar to a fee), whereas full-time teachers receive a salary. In both cases, the compensation is a market (or institutional) valuation of the teacher’s labor output.

Institutions

The absence of an explicit fee for some services does not raise any particularly unique problems. Rather, the wage (plus benefits and overhead) is effectively the price an institution (such as a government agency or a nonprofit) decides it is worth paying to employ that individual. You may disagree with how that institution makes its decisions, but that is an issue of institutional quality, not of service valuation as such.

If an institution is poor at valuing services, there is no particular reason to assume it would be any better at valuing goods. Little about the nature of a service, compared to a physical good, makes its valuation inherently more or less prone to miscalculation.

Baumol’s Cost Disease and Other Service-Specific Idiosyncrasies

Economist William Baumol observed that wages in labor-intensive sectors such as education and the arts tend to rise even when productivity in these fields remains relatively stagnant. This is because of overall wage increases in the broader economy. As productivity grows in other sectors, the opportunity cost of working in, say, cosmetology also rises, which pushes up wages for people in this industry.

At first glance, this may seem problematic because we give up more for a haircut or a concert performance even when the product hasn’t changed much. But from an opportunity-cost perspective, this outcome is logical. If other sectors are more productive, the cost of labor has gone up, so services that require one-on-one labor become relatively more expensive over time. This is not an economic anomaly but rather a natural reflection of rising opportunity costs in a dynamic and growing economy.

Additionally, some services are difficult to trade internationally (like haircuts or in-person teaching), which can introduce further quirks in how they are priced. Non-tradable services may indeed create market failures, such as mispricing of exchange rates between countries’ currencies. Yet there are also common market failures associated with the production of physical goods. For example, environmental externalities stemming from factory production. There is no obvious reason why services are systematically more prone to market failures than physical goods.

Can We Trust Our Statistics?

Services are generally subject to the same market forces and opportunity-cost logic as any other economic activity. While services may have particular peculiarities like Baumol’s cost disease or non-tradability, these do not imply that the value of services in GDP is systematically mismeasured. GDP captures the market valuation of output, not the ideal shadow price in a perfect market.

If we want welfare analysis to accurately capture social value, cost-benefit analysis is better-suited for that purpose than GDP. The latter would require updating all market prices and converting them to shadow prices. Unlike GDP, CBA’s intended purpose is to assign shadow prices to both goods and services. However, in CBA the same challenges arise in valuing externalities and dealing with institutional failings regardless of whether the product is tangible or intangible. That’s because most market prices are distorted by market failures.

Some Concluding Thoughts

When someone says “services are valued at cost,” they seem to be pointing to the distinction between accounting cost (wages) and economic cost (opportunity cost to society). In practice, market prices paid depart from true social value of products for many reasons, including externalities, taxes, subsidies, or simply imperfect institutions.

There is, however, nothing uniquely problematic about services. Perhaps society wants more investment and many services represent pure consumption. But the same concern can be expressed about consumption of physical goods. Like services, the opportunity cost of physical consumption likely rises with time for Baumol-like reasons.

Some might argue market failures occur more frequently for services because governments often purchase services for third parties (e.g., education or healthcare). But similar issues arise when governments buy physical goods for foreign aid or when private entities purchase tangible and intangible goods for others (charity, gifts, insurance claims, etc.).

In the end, the list of issues that create gaps between actual prices paid and ideal shadow prices is almost endless, affecting goods and services alike. One hopes that over time, prices and wages tend to move in the direction of reflecting social opportunity costs. That may not always be the case, as Baumol’s cost disease implies may sometimes be the case. But whether we are discussing a teacher’s labor, a carton of orange juice, or the services provided by your smartphone’s latest app, the same economic principles apply.