The expansion of broadband internet in Norway resulted in firms exhibiting increased productivity, greater wage differences between individuals with high and low levels of education, and an increase in the negative impact of distance on international trade. These are the results in one of the few studies having successfully shown the causal effects of digital infrastructure on the economy. This study is presented in the new SNS report “Economic impact of investments in digital infrastructure” by economist Anders Åkerman.
Despite the fact that there are numerous indications that advances in information technology have an impact on economic development, it has been notoriously difficult to measure the causal effects of, for example, internet access. Hence, this study serves as an important contribution when it comes to gaining a better understanding of the impact of digital infrastructure on productivity, trade, the labor market as well as the magnitude of these effects. The gradual expansion of internet broadband in Norway in 2000–2008 was in the study used as a so-called natural experiment.
The results show that the Norwegian internet broadband expansion resulted in firms on average exhibiting a four percent increase in productivity. In addition, the labor market for university graduates clearly improved compared to the labor market for people with a lower level of education, both in terms of wage increases and employment rates. One explanation for this is that the tasks facilitated the most by the internet were mainly carried out by university graduates. As a result, their productivity increased, while it remained the same for non-university graduates. This, in turn, is reflected in how wages developed. However, it is not obvious that technological developments automatically favor workers with a high level of education.
“The tasks that may be performed more efficiently thanks to new technology are not static but change over time,” says Anders Åkerman, Associate Professor at the Department of Economics, Stockholm University.
The report also shows that internet access increased the significance of distance with regards to trade. All else being equal, trade with neighboring countries increased in comparison to more distant countries. An explanation for this is that increased information resulted in increased competition, which, in turn, increased the significance of transport costs. These effects were mainly seen with regards to goods offered in many variations, such as cars and computers. Comparing these types of goods requires a great deal of information, something simplified by the internet.
“Our results indicate that digitalization does not necessarily lead to a ‘flat’ world where geographical distances no longer matter. On the contrary, there is now talk of increasing regionalization, and goods produced closer to the consumer play an increasing role”, says Anders Åkerman.
Understanding the effects of the first internet broadband expansion is also important when it comes to the future; for example, with regards to the 5G expansion and the increasing use of artificial intelligence and machine learning.
“Many people believe that these technologies will fundamentally change society. It is thus important to create an early idea of who is likely to benefit and who may encounter problems”, says Anders Åkerman.
This report is published in the framework of the SNS research project Sustainable Urban and Rural Planning.
About the report
The report is based on two studies: one on productivity and the labor market, carried out by Anders Åkerman together with Assistant Professor Ingvil Gaarder and Professor Magne Mogstad, and one on trade carried out together with Professor Edwin Leuven and Professor Magne Mogstad.
Measuring the impact of infrastructure or technological advances is often difficult. There is rarely a clear control group or counterfactual outcome to compare with. In these studies, the researchers use the gradual expansion of internet broadband in Norway in 2000–2008 as a so-called natural experiment.
In these kinds of studies, the researcher uses a variation similar to that in a controlled experiment. However, the researcher does not control the variation; instead, it may be the result of, for instance, a rule, reform or event creating a more or less random variation in the treatment. Due to this random nature, it is possible to compare the outcomes for those who received the treatment and those who did not. The difference in outcome may then be interpreted as a causal effect of the treatment. Since the timing of the internet broadband expansion was more or less random for a municipality’s firms and residents, the outcomes may be compared between municipalities that had internet broadband at a given point in time and municipalities that did not. How these differences in outcomes change over time as access to the internet increases may be interpreted as a causal effect of internet broadband access.
The report used detailed data on individuals and firms, including information on the production and international trade of all firms, as well as information on individual wages, employment status, and education. Information on firms’ internet access has also been used.