What Swiss highways can tell us about inequality

Transportation infrastructure shapes the spatial economy in fundamental ways, with effects on regional disparities that are ambiguous at best (e.g. Baum-Snow et al. 2020, Faber 2014, Koster et al. 2021). Commuting allows people to access labour markets and amenities that are distant from their homes and workers to access housing markets that are remote from their workplace.¹ Therefore, by facilitating commuting, highways make connected municipalities more attractive.

In a new paper with Stephan Fretz (Fretz et al. forthcoming), we study the consequences of the development of a major transportation infrastructure over a long period – the development of the Swiss highway network from 1960 to 2010 – on the sorting of residents and workers with heterogeneous incomes and skills. To the best of our knowledge, our paper and the contemporaneous work by Tsivanidis (2018, 2019) are the first to examine the effect of transportation infrastructure on the spatial allocation of heterogeneous workers.²

Switzerland is a small country with comprehensive railway, highway and road networks, and a population heavily concentrated on the hilly Swiss Plateau, which covers only 30% of the territory. Hence, many municipalities are within commuting distance of at least one town.

The Swiss highway network as of 2010 is represented in black in Figure 1; the four largest cities (Zurich, Geneva, Basel, Lausanne) and the largest Italian-speaking city (Lugano) are also shown. Urban municipalities appear in dark grey, and rural municipalities that are connected appear in light grey. The yellow dots indicate the location of highway ramps opened by 1970 (panel a) and 2010 (panel b).

Figure 1 The Swiss highway network

(a) Highway access opened by 1970

(b) Highway access opened by 2010

Notes: Black lines are the Swiss highway network, as of 2010. Dark grey areas are urban municipalities; light grey areas are rural municipalities that are connected. Yellow dots indicate the location of highway entrance/exit ramps opened by 1970 (panel a) and 2010 (panel b).

The country's demographics shifted radically over the period under study. First, the overall Swiss population grew from about 5.3 million in 1960 to 7.8 million in 2010 (and 8.7 million today).

Second, this remarkable growth hides substantial spatial heterogeneity:

Figure 2 illustrates points 1, 3, and 4. Municipalities whose population grew faster than the overall Swiss population appear in green; municipalities that lost relative to the country average, in red. Urban municipalities are represented in dark colours; rural municipalities that are connected by 2010 in intermediate colours; and rural municipalities that remain unconnected in 2010 in light colours. The eyeball correlation between proximity to the highway and likelihood of appearing in green is in the eye of the beholder; the actual correlation is positive and strong.³

Figure 2 Relative population gains by municipality, 1960–2010

Notes: Green areas are municipalities whose population grew faster than the overall Swiss population; red areas are municipalities that lost population relative to the country average. Dark colours are urban municipalities; intermediate colours are rural municipalities that are connected by 2010; light colours are rural municipalities that remain unconnected in 2010.

In our paper, we document that the development of the highway network played a substantial role in points 1 to 5. In particular, we find that the presence of a highway entrance/exit ramp within 10km of a municipality caused a long-term 24% increase in the share of top-income taxpayers living in the area and an 8% decrease in the share of below-median income earners.

Since low-income earners were initially over-represented in municipalities that became connected relative to the country average, highways reduced segregation in Swiss municipalities.

Understanding the spatial and economic consequences of large-scale transportation infrastructures is important for several reasons. First, access to markets and proximity to workers and jobs are prominent criteria in the location decisions of firms and households. Therefore, transportation infrastructures are an important determinant of individual welfare and of regional disparities.

Second, the location of airports and the design of rail, road, and highway networks influence land-use patterns as much as ‘first nature’ geography: highways have been found to increase the size of cities (Duranton and Turner 2012), cause suburbanisation (Baum-Snow 2007, Baum-Snow et al. 2017, Brinkman and Lin 2020, Garcia-López and Viladecans-Marsal 2016), affect the product mix of cities (Duranton et al. 2014), and increase regional disparities (Baum-Snow et al. 2020, Faber 2014).

Third, at around 5% of GDP, the amounts of money involved in transportation infrastructure dwarf those of most other investment programmes (Redding and Turner 2015) but may also bring large-scale economic benefits (Allen and Arkolakis 2014, Donaldson 2018, Donaldson and Hornbeck 2016).

Furthermore, as we document below, highway access leads to worker sorting along skills and incomes, which is likely to have implications for voting and, in federal countries such as Switzerland that grant large budget autonomy to their municipalities, on tax competition (Eugster and Parchet 2019, Parchet 2019).

We start by developing a parsimonious spatial equilibrium model featuring costly commuting. The model features a labour force with heterogeneous skills and identical, non-homothetic preferences over housing and other goods and services, and idiosyncratic location and commuting preferences.

Guided by Swiss data, we assume that preferences are such that richer workers spend a lower fraction of their income on housing than poorer workers (i.e. housing is a necessity), and they are more likely to commute by car (the more expensive commuting mode).

Connection to the highway network increases the commuting access of a municipality, which leads to two effects in the model. First, it attracts residents, which raises local demand for housing and increases local housing prices. This effect disproportionately hurts low-income earners because they spend a higher-than-average fraction of earnings on housing. Second, improving commuting access by car disproportionately benefits the well off because they are more likely to own and commute by car.

Both mechanisms yield the same qualitative outcome: a newly connected municipality becomes especially attractive to high-income, high-skilled residents. The second mechanism also makes such a municipality relatively attractive to high-skilled workers. As a result, its skill and earnings distributions shift up.

We then empirically investigate the effect of improved accessibility on income distribution at the local level, exploiting variation in the commuting access of municipalities over time from the construction of the Swiss highway network.

Switzerland provides an ideal setting for this study as the Swiss highway network was planned in 1960 by the federal parliament to connect Switzerland’s largest cities, but was only gradually constructed over the decades that followed. From the perspective of a non-urban municipality, the opening date of a new highway section in its vicinity is close to random and exogenous to the initial path of its local economic development.

Merging several data sets, we exploit this variation over time to identify the effect of the opening of a new highway access point on the total number of taxpayers (resident households), the share of taxpayers in different income categories, as well as employment and commuting by education levels.⁴

We also account for a rich combination of unobserved factors by including municipality-specific fixed effects and linear time trends, as well as year dummies, thereby mitigating sources of omitted variable bias. We also explicitly account for the presence of railways.

We find that the number of taxpayers and the share of top-income taxpayers rise in non-urban municipalities located within 10km of newly opened highway access points. We also find that segregation by income (as measured by the Theil entropy index) falls in connected municipalities.

Specifically, the total number of taxpayers increases by 14%, the share of low-income taxpayers (defined as taxpayers with earnings below the median) decreases by 8%, and the share of top-income taxpayers (taxpayers in the top decile of the earnings distribution) increases by 24%. These causal estimates translate into a 5% increase in the number of low-income taxpayers and a 42% increase in the number of top-income taxpayers.

Finally, we estimate the relative long-run welfare effects of the expansion of the Swiss highway network. We find that, over the years 1950–2010, the wellbeing of residents in non-urban municipalities with a highway connection in 2010 increases relative to that of residents in non-urban municipalities that were still unconnected by 2010. This relative welfare gain increases monotonically with the income quintile, from only 2% for the below-median income group to 12% for the top-10% income group.

Our model relates these differences in welfare changes between income groups to housing being a necessity and car ownership and car use being a luxury good. Specifically, access to highways only benefits car users and the intensity of car use increases with income; access also increases the attractiveness of a municipality, which raises housing prices and rents, and high-income earners spend a lower share of earnings on housing.

Car use is a luxury; housing is a necessity. Thus, the benefits of highway access disproportionately accrue to the well off, and its costs disproportionately hurt low-income earners. Municipalities that get access to the highway network become relatively attractive to the best off and unattractive to the least well off.

The expansion of highway networks thus shapes the economic geography in a fundamental way: it alters the income distribution of municipalities and hence regional disparities. Using the development of the Swiss highway network over fifty years, this column provides conclusive evidence of this deep economic mechanism. Especially for Swiss municipalities, who have considerable fiscal authority and are engaged in tax competition, highways have thus a potentially large impact that merits further research.

Allen, T, and C Arkolakis (2014), “Trade and topography of the spatial economy”, Quarterly Journal of Economics 129(3): 1085–140.

Baum-Snow, N (2007), “Did highways cause suburbanization?”, The Quarterly Journal of Economics 122(2): 775–805.

Baum-Snow, N, L Brandt, J V Henderson, M A Turner and Q Zhang (2017), “Roads, railroads, and decentralization of Chinese cities”, Review of Economics and Statistics 99(3): 435–48.

Baum-Snow, N, J V Henderson, M Turner, Q Zhang and L Brandt (2020), “Does investment in national highways help or hurt hinterland city growth?”, Journal of Urban Economics 115: 103–24.

Brinkman, J, and J Lin (2020) “Freeway revolts! The quality of life effects of highways”, Federal Reserve Bank of Philadelphia Working Paper 19-29.

Donaldson, D (2018), “Railroads of the Raj: Estimating the impact of transportation infrastructure”, American Economic Review 108(4-5): 899–934A.

Donaldson, D, and R Hornbeck (2016), “Railroads and American economic growth: A ‘market access’ approach”, The Quarterly Journal of Economics 131(2): 799–858.

Duranton, G, P Morrow and M Turner (2014), “Roads and trade: Evidence from the US”, The Review of Economic Studies 81(2): 681–724.

Duranton, G, and M Turner (2012), “Urban growth and transportation”, The Review of Economic Studies 79(4): 1407–40.

Eugster, B, and R Parchet (2019), “Culture and taxes”, Journal of Political Economy 127(1): 296–336.

Faber, B (2014), “Trade integration, market size, and industrialization: Evidence from China’s national trunk highway system”, Review of Economic Studies 81: 1046–70.

Fretz, S, R Parchet, and F Robert-Nicoud (2021), “Highways, market access, and spatial sorting”, Economic Journal, forthcoming.

Garcia-López, P, and E Viladecans-Marsal (2016), “Express delivery to the suburbs. Transport infrastructure and European cities”, CESifo Working Paper 5699.

Hayakawa, K, H Koster, T Tabuchi, and J F Thisse (2021), “How high-speed rail changes the spatial distribution of economic activity: Evidence from Japan’s Shinkansen”, VoxEU.org, 8 April.

Koster, H, T Tabuchi and J F Thisse (2021), “High-speed rail may hurt intermediate places: The role of long-haul economies”, VoxEU.org, 9 May.

Monte, F, S Redding, and E Rossi-Hansberg (2015), “Commuting, migration and local employment elasticities”, VoxEU.org, 10 December.

Parchet, R (2019), “Are local tax rates strategic complements or strategic substitutes?”, American Economic Journal: Economic Policy 11(2): 189–224.

Redding, S, and M Turner (2015), “Transportation costs and the spatial organization of economic activity”, in G Duranton, J V Handerson and W C Strange (eds.), Handbook of Regional and Urban Economics, Elsevier, 1339–97.

Tsivanidis, N (2019), “Evaluating the impact of urban transit infrastructure: Evidence from Bogotá’s TransMilenio”, mimeo, UC Berkeley.

Tsivanidis, N (2018), “The equitable benefits of Colombia’s bus rapid transit system”, VoxDev, 1 May.

1 On the importance of commuting for employment responses to local economic shocks, see e.g. Monte et al. 2015. On the importance of commuting for the spatial distribution of economic activity, see e.g. Hayakawa et al. 2021.

2 Tsivanidis (2019) evaluates the impact of Bogotá’s TransMilenio (the largest rapid bus transit system at the time) on the modal choices of Bogotá’s residents and on the composition of its neighbourhoods.

3 The large alpine municipalities in the south and west of the country are at very high altitudes and are sparsely inhabited.

4 The federal authorities allowed to make a substantial subset of this data publicly available.