This is the final report in a knowledge project where we have studied productivity differences between regions and how commuting can help spread the productivity gains of larger cities to surrounding smaller areas.

High wages in large cities – selection or agglomeration economies?

International research indicates that wage levels are higher in large cities and regions. In this context, measures such as nominal hourly wages or annual gross earnings are used as an indicator of the productivity benefits of different locations. The argument is that if firms in large cities and regions are to be able to afford to pay high wages, these places must offer particular productivity advantages. Otherwise, firms operating in these environments would be pushed out of the market by competitors located in places with lower wage costs.
Sweden is no exception to the international picture. On average, the wages of people working in the Stockholm region are approximately 35 percent higher than those of people working in Sweden’s smallest regions.

What is the explanation for the fact that wage levels tend to increase with the size of cities and regions, i.e. that there seems to be a so-called urban wage premium? One possible interpretation is that it depends on selection. According to this view, larger cities/regions succeed in attracting individuals and businesses that are more productive and thus firms operating in these locations can afford to pay higher wages. Another possible explanation is that the economic environment in larger cities/regions is particularly favourable and in various ways contributes to increasing the productivity of the individuals and firms that operate there. It is the latter idea to which the concept of agglomeration economies refers.
The fact that agglomeration economies tend to arise in large cities/regions is usually explained by three basic theoretical mechanisms:

• One of these is that larger cities create a basis for sharing important resources characterised by indivisibilities and high investment costs, such as various types of infrastructure.
• Another important mechanism is that it is easier to achieve high quality in the matching between work tasks and skills in larger labour markets.
• A third mechanism has to do with learning and takes into account that larger cities have particularly good conditions for the creation and diffusion of knowledge.
  The education

The education gap between large and small regions is growing, and young people’s migration patterns are widening this divide

Our analyses show a clear polarisation of human capital assets between local labour markets over the past three decades. The supply of high skilled workers is significantly better in Sweden’s larger regions, and the share of high skilled workers has increased most in regions with high initial shares. The migration patterns of the highly educated population reinforce the education gap between large and small regions. The metropolitan regions, and the Stockholm region in particular, receive significant net inflows of highly educated young people. At the same time, the smaller regions show significant migration losses. But even within the group of highly educated people, there is a clear regional sorting of human capital, where metropolitan regions attract individuals who are likely to be particularly ambitious and productive.

In recent years, several studies have found that there will be a major shortage of labour in the coming decades, especially in smaller municipalities and regions. This applies both in business sector and in the public sector. Not least, it is expected to be difficult to recruit highly qualified workers to various specialist positions and management functions in the municipal sector.

Both more general and specific measures have been proposed to facilitate the supply of skills in different types of region. One potentially interesting proposal, which specifically focuses on the supply of highly skilled workers, is to write off student debt for people who live and work in areas with relatively poorer recruitment opportunities. There are positive experiences from such a system in Norway. The Parliamentary Rural Committee (Parlamentariska landsbygdskommittén) has proposed that the government investigate the possibility of implementing a similar system in Sweden (SOU 2017:1). The Municipal Inquiry (Kommunutredningen) has also recently recommended that the government investigate the possibilities to write off student debts (SOU 2020: 8).

Regional income disparities have narrowed since the turn of the millennium, but seem to increase in times of crisis

Although highly skilled workers are increasingly seeking out metropolitan areas, we see no sign of growing regional disparities in per capita income. However, the recovery period following the financial crisis of the 1990s is an exception. During that period, income inequality increased considerably. Similar trends have been observed during previous economic crises. However, since the turn of the millennium, and especially in the most recent years (2015-2018), regional disparities in per capita income have decreased significantly.

The development in Sweden since the turn of the millennium differ from the trend towards increasing regional income inequality observed in the US and many countries within the EU. This may indicate that the Swedish model of welfare state, with a relatively large redistribution of economic resources between individuals and regions, is more effective in equalising regional inequalities than other variants of welfare models. At the same time, however, we note that the long-term trend towards decreasing regional income inequality in Sweden coincides with significant net out-migration and a shrinking population in many of the country's smaller regions. With declining population, municipal and regional tax bases also erodes. At the same time, different types of additional structural costs can arise.

An important question in this context is whether the existing design of the municipal financial equalisation system takes sufficient account of the economic impact of the sharp population decline that many of the country's smaller regions have gone through and continue to experience, and whether the equalisation system thus meets the objective of guaranteeing municipalities equivalent financial conditions. The recently presented
Municipal Inquiry (Kommunutredningen) has proposed that the government appoint an inquiry with the task of conducting a broad review of the municipal financial equalisation system (SOU 2020:8).

Migration from smaller to larger regions and especially to the Stockholm region yields positive wage effects

Using a matching method, we identify “stayers” and “movers” who are comparable based on a variety of characteristics and use the estimated effects of migration on individuals’ gross wages as an indicator of regional differences in productivity. We find that migration upwards in the regional hierarchy results in significant positive effects on wages. Migration downwards in the hierarchy is generally associated with negative effects on wages. The wage premium of moving to the Stockholm region is particularly high. We interpret these results as an indication that large regions display productivity advantages compared to smaller regions. There are several possible theoretical mechanisms that can explain why large regions seem to have particular productivity advantages. These include more effective job matching in large regions and the fact that large regions offer better conditions for knowledge diffusion and learning.

More recently, several actors have signalled that restrictions in the housing market constitute an obstacle to migration to growth regions. This creates problems for those who wish to migrate to these areas. It also makes it difficult for firms that want to recruit staff. In the end, this entails a cost to society in terms of loss of potential productivity gains.

Commuting help spread the productivity gains of larger cities to surrounding smaller areas nearby

The extent to which workers commute between Swedish municipalities has increased dramatically in recent decades. Today, more than one in three employed people commute to a job outside the municipality of residence. Most people commute from a smaller municipality to a larger one.

In our report, we study commuting to five categories of urban municipalities. The inclination to commute increases with the size of the destination municipality, and diminishes with greater distances.

Using a matching method, we identify commuters and non-commuters who are comparable based on a variety of characteristics and use the estimated effects of commuting on individuals’ gross wages as an indicator of productivity advantages in urban municipalities. We find that there are positive wage premiums for starting to commute into all five categories of urban municipalities. The premium tends to increase with the size of the destination municipality, and the wage effect is greatest for commuting to Stockholm.
This result may be interpreted as an indication that municipalities at the top of the urban hierarchy show the greatest productivity advantages. There are also signs of lesser productivity advantages in medium and small urban municipalities. When it comes to commuting to Stockholm, we find that people with a high level of education or experience in a highly qualified occupation receive a particularly high wage premium.
This suggests that it is easier for highly qualified workers to find jobs that match their qualifications in Stockholm's large and diversified labour market.

We also note that the wage premium tends to be higher for individuals who start commuting from more peripherally located municipalities into an urban centre. At the same time, the volume of commuting from these municipalities into city centres is considerably smaller than for the more nearby municipalities. This means that in the more remote municipalities, increased commuting into urban centres entails a great potential for productivity gains.

Commuting costs are currently subsidised, including via a deduction on taxable income for expenses related to travel to and from work. In the current system there are requirements for both minimum distance and time gain (for certain means of transport) in order to be entitled to make a travel deduction. The Travel Deduction Committee (Reseavdragskommittén) recently proposed that the travel deduction should be abolished in its current form and replaced by a distance-based and means-neutral tax reduction (SOU 2019:36). According to the proposal, the tax reduction is to be granted for that part of the distance between one’s residence and the workplace that exceeds 30 kilometres (one-way), up to a maximum of 80 kilometres. A certain additional tax reduction is also proposed in cases where public transport is inadequate. There is reason to ponder and carefully consider the maximum distance limit in the proposal. As we have seen, the extent of commuting over greater distances is admittedly relatively limited. At the same time, our results indicate that the potential benefits of increased commuting from more peripherally located municipalities can be significant.

“Regional enlargement” via teleworking instead of daily travel?

It is well established in previous research that commuting in its traditional form – daily travel between home and work – is associated with various types of costs and problems. These include negative consequences for well-being, health, gender equality and the environment.

Long before internet and broadband became an obvious part of everyday life, there were visions of a different form of “regional enlargement” that was not based on daily travel. High expectations were attached to the then emerging technology for telecommunications, which it was hoped would pave the way for a more flexible working life, with a greater element of teleworking. But the new technology did not bring with it the boost for towns and communities outside the big cities that many hoped for. The current coronavirus pandemic can be seen as an involuntary, full-scale experiment on the true potential of teleworking.

High-speed broadband in all parts of the country is an important infrastructure prerequisite for teleworking. The government’s broadband strategy A Completely Connected Sweden by 2025 (Regeringskansliet 2016) establishes the goal that all households and companies should have access to broadband with a speed of at least 100 Mbps by 2020. In the longer term, ambitions are higher. The goal is that by 2025, 98 percent of all households and businesses should have access to broadband with a speed of at lest 1 Gbps. In a recent follow-up to the government’s broadband strategy, the Swedish Post and Telecom Authority (Post- och telestyrelsen 2020) concludes that the short-term target will not be achieved before the end of 2020, and that the long-term goal will not be fully achieved.

From a localisation and agglomeration perspective, there are several interesting aspects to the possible effects of increased teleworking. How are the downtown areas of major cities affected by a possible reduction in demand for office space and lower turnover for restaurants and shops? Will an increase in teleworking lead to the vision of a new form of regional enlargement, which to a greater extent benefits development in sparsely populated and rural areas, being realized this time? These are examples of interesting questions that necessitate new knowledge and research.