The following is an article published by "Guangming Daily" on December 19, 2017. The author is Tan Hongbo. The original title is "Can artificial intelligence cure Baumol's disease?"
Artificial Intelligence is currently one of the hottest topics in the fields of science and technology and economy. The reason why it is hot is mainly because it is or will soon bring a disruptive revolution to the economy and society. Discussing the development of artificial intelligence from an economic perspective mainly starts from the replacement of labor by artificial intelligence, which will inevitably involve the famous "Baumer's disease" problem.
American economist Baumol constructed a two-sector non-equilibrium growth model in 1967, which successfully explained the changes in the industrial structure of major economies during most of the 20th century and the economic growth trends. Baumol divided the macroeconomy into progressive sectors with positive labor productivity growth rates and stagnant sectors with no labor productivity growth rate, and based on several key assumptions: over time, the unit product of the progressive sectors Costs will remain constant (here labor costs), while the unit product costs of the stagnant sector will continue to rise. Therefore, consumer demand for the products of the stagnant sector will be, if not perfectly price inelastic, so the rising unit product costs of the stagnant sector will Consumers will be prompted to reduce their demand for the sector's products, with the result that the stagnant sector will shrink and eventually disappear.
Assuming that the demand for products in the stagnant sector is completely inelastic, then although the unit product cost of the stagnant sector continues to rise, there will still be a continuous inflow of labor into this sector, so that this sector will not only not shrink but will also Gradually absorbing a large number of labor forces, precisely because the labor force continues to transfer from progressive sectors to stagnant sectors, the economic growth rate of the entire country will gradually become zero. This is the famous Baumol cost disease and growth disease, or Baumol disease for short.
Baumol also pointed out that the progressive sector mainly refers to the manufacturing industry, while the stagnant sector refers to the service industry, including education, municipal services, performing arts, catering, entertainment and leisure, etc. Baumol pointed out as an example, In the performing arts market, it took four people to play Mozart's quartet 300 years ago, but it still takes four people to play the same piece of music 300 years later. Labor productivity has not changed.
The theoretical reasoning of Baumol's disease is simple and clear, and it can also explain the high cost of the service industry in market economy countries. However, the assumptions of Baumol's 1967 theoretical model were too simple. For example, there were only two sectors, the service industry was only the final product, and an open economy was not considered. Therefore, some subsequent scholars conducted research on Baumol's disease from both theoretical and empirical levels. expanded and tested. Theoretical research is mainly an expansion of the Baumol disease model. Based on previous research, Baumol (1985) expanded the model into a three-sector model in order to overcome the limitation of treating the entire service industry as a stagnant sector. In addition to In addition to the progress sector and the stagnation sector, the progressive stagnation sector is also introduced.
In addition, Ngai and Pissarides (2007) theoretically extended Baumol's disease more generally. Based on Baumol's theory, they assumed that there are n departments, and each department has Different labor productivity growth rates. Baumol (1985), Ngai and Pissarides (2007) obtained basically the same conclusion as Baumol (1967) after extending the model. Oulton (2001) assumes that the service industry is no longer a final product sector, but a sector that provides intermediate inputs for the secondary industry. From this, he draws the same conclusion as Baumol in terms of changes in industrial structure - the labor force is still continuously changing from The secondary industry flows to the service industry, but at the same time, it reaches a different conclusion from Baumol in terms of economic growth—the economy will not stagnate in the end.
In terms of empirical testing, one type of research believes that Baumol's disease exists in major developed countries. Baumol (1985) used historical data from 1947 to 1976 in the United States to empirically test and found that during this period, the progressive sectors and stagnation in the United States The sectors' shares of real output are fairly stable, so as relative prices rise, the share of spending and labor in stagnant sectors, services, rises sharply.
Nordhaus (2008) and Hartwing (2011) used data from the United States and the European Union from 1948 to 2001 respectively and found that Baumol's cost disease and growth disease existed in both the United States and the European Union.
Another type of research found that Baumol's cost disease and growth disease do not exist in developed countries such as the United States. Triplett and Bosworth (2004) used U.S. data from 1995 to 2001 to conduct an empirical study and found that the average labor productivity in the service production industry increased by 2.3%, while the growth rate in the tangible product production industry was only 1.8%. They believed that there was no Bao in the United States. Moore's disease. Timmer et al. (2007) used the EU KLEMS database to conduct empirical analysis and found that labor productivity in the service industry has accelerated in Greece, Ireland and the Netherlands since 1995. Their empirical results also confirmed that Baumol's disease does not exist in these countries.
From the two-sector model originally proposed by Baumol (1967) to the three-sector model, and then to the general n-sector model, the most critical assumption in the above research is the rate of progress in labor productivity between different sectors. This is also the main focus of debate among different scholars. However, with the development of science and technology, large-scale differentiation has gradually begun within the service industry. Some service industries have experienced rapid growth in labor productivity due to the large-scale application of ICT technology. The growth rate even exceeds the growth rate of manufacturing labor productivity as a progressive sector, such as finance, telecommunications, commercial circulation and other industries.
At the same time, some service industries are still as described by Baumol, with slow progress in labor productivity and rising costs, such as domestic services, beauty salons, medical diagnosis, education, etc. According to Baumol's example, in the field of educational services, it took one semester to teach "Principles of Economics" 30 years ago, but it will still take one semester to teach the same course 30 years later. Thirty years ago, the average men's haircut took a barber 20 minutes. 30 years later, the same haircut service still requires a barber's 20 minutes. It can be seen that the labor productivity of these traditional service industries has progressed very slowly, so these traditional service industries still belong to the typical stagnant sectors described by Baumol. Of course, the scope of these traditional sectors may be narrowed with the promotion of technology.
Whether it was the era when Baumol proposed the theory or in recent decades, Baumol's disease has appeared more or less in many market economy countries, and is particularly prominent in developed countries. For example, it manifests as The repair cost of many appliances or equipment is higher than the replacement cost because repair is a traditional service industry, while producing a new product is a manufacturing industry. Another example is that ordinary cleaners can generally afford family cars more easily, because the work of cleaners belongs to the traditional service industry, while car production belongs to manufacturing.
However, the author believes that the emergence and development of artificial intelligence will completely change the basis for the existence of Baumol's disease.
In terms of educational services, in the future, it will no longer be teachers, but artificial intelligence teachers. This kind of artificial intelligence education can provide teaching services anytime and anywhere according to the personalities of different students. Maybe it will not be long before , like the article the author is writing, can be completed by artificial intelligence, and the speed is faster and the analysis is more in-depth. Various services that people go to the city hall will also be replaced by artificial intelligence. Therefore, in the future, bank tellers, restaurants Professional positions such as counters, receptionists, shopping mall sales and service personnel, teachers, doctors, financial analysts, accountants, and lawyers may all be replaced by artificial intelligence on a large scale.
It can be seen from these examples that artificial intelligence will replace not only modern service industries with high labor productivity growth rates, such as finance, accounting, etc., but also the labor productivity that Baumol referred to back then. Industries with slow or stagnant growth, such as education, healthcare, catering, performing arts, etc. Therefore, the emergence of artificial intelligence has caused a sharp increase in labor productivity in all industries, and many industries will no longer even have a labor force. In other words, in the era of artificial intelligence, there is no industry that is what Baumol calls a stagnation sector or a progressive stagnation sector. This Completely overturned the basis of Baumol's disease.
After artificial intelligence replaces labor on a large scale, as the labor productivity of all service and manufacturing industries continues to grow rapidly, according to Baumol’s reasoning, the unit cost of products and services will continue to decline. The economy will no longer stagnate, but will continue to grow, so the development and widespread application of artificial intelligence will cure Baumol's disease.
(Author: Tan Hongbo, deputy director of the Economics Department of Yangzhou University Business School, this article is a phased result of the National Social Science Fund Project [16BJY129])