The service sector does not consist of a homogeneous group of services. The industries within the service sector are too heterogeneous for a common frame of analysis. In this article we are using a classification scheme proposed by Baumol (1984) with some modifications. The services can be classified into four categories:
1. Stagnant personal services
2. Substitutable personal services
3. Progressive services
4. Explosive services.
These services frequently require direct contact between the customers and the service provider. Some examples are haircutting, live artistic performance, psychiatric counseling, and teaching. Since the quality of such a service is highly correlated with labor time, it is difficult to realize significant productivity gains for these services without an appreciable reduction in quality. It seems evident, for instance, that the amount of time required for a haircut cannot be decreased substantially without some drastic implications. These services offer low innovation potential and are difficult to standardize.
These services also require direct personal contact, and they have characteristics similar to stagnant personal services. However it is possible to substitute for these services with technological or other alternatives. An example would be the services of a guard that can be replaced by electronic surveillance systems. In the twentieth century, we have seen ovens, washers and dryers and other household appliances substituted for personal servants (maids, cooks etc). Some types of teaching such as real estate licensing course work, are now conducted using cassettes, tapes, and videos, and the demand for the live instructors as decreased correspondingly.
A great leap in productivity in substitutable personal services is provided by technological innovation.
These services have two components. One component requires little labor, and considerable cost reductions are possible with it. The second component is highly labor intensive and is much like stagnant personal services. An example is computation services.
In a simplified aggregate view, computation services can be conceptualized as consisting of hardware and software. The cost of hardware per computation has declined steadily; conversely, the cost of software has risen. This is because software is produced by human labor and offers limited productivity growth.
Another example is television broadcasting, where the two components are the transmission and the production of a program. Transmission costs have steadily decreased with advances in the electronics and space industries. However, the production of a television program is highly labor intensive, and consequently, the associated cots continue to increase. Research and development can also be thought of as consisting of two dichotomous components: equipment and human thought.
Progressive services can exhibit phenomenal productivity growth and cost reductions initially. This is due to the relatively important contribution of the first technology intensive component. For example, computer hardware contributes significantly to computation costs; thus, decreases in hardware costs per computation lead to overall cost reductions. Since the costs per unit of output for the second, labor intensive component are increasing, the decline in total cost cannot be sustained for long periods. In this sense, productivity growth is self-extinguishing. This happens because, in due course of time, the relative contribution of the second component exceeds that of the first component. The stagnant nature of the second component dampens productivity growth.
A numerical example illustrates the cost per unit behavior of progressive services. Suppose the share of component 1 is 80% in the beginning and that of component 2 is 20%. Further, suppose that the cost of the first component is declining at the rate of 30% per year and the cost of the second component is rising at the rate of 10% per year. For an output of service that costs $100 in year zero (the initial period), the first component costs $80 and the second costs $20. At the end of the first year, the cost of the first component has declined to $80 â€“ (0.3 x $80) = $56, and the cost of the second component has risen to $20 + (0.1 x $20) = $22. Thus, the total cost of the same unit of service has decreased from $100 to $56 + $22 = $78. In similar manner, we can compute the cost of the two components for year two and so on.
An important feature of the progressive service is that the faster the initial decline in total cost per unit of service, the more rapidly cost reduction ceases. This is because the relative share of the first component responsible for cost reductions shrinks rapidly and the service begins to acquire the characteristics of the second component.