Fifty years of urban accessibility: The impact of the urban railway network on the land gradient in Berlin 1890–1936

https://doi.org/10.1016/j.regsciurbeco.2010.10.001Get rights and content

Abstract

As the first to use an archival data set on historical land values in Berlin, Germany, from 1890 to 1936, we investigate the impact of the rapid transport system on urban decentralization, using comparative statics of classical rent theory as a benchmark. We find that the monocentric model performs well over the entire period studied, revealing gradients that – although diminishing over time – turn out to be relatively steep in international comparison. Travel time to CBD measures incorporating the rapid transport network, however, clearly outperform traditional distance to CBD measures in terms of explanatory power. The evolution of the rapid transit network, and the subsequent changes in travel times to the CBD, explain almost three quarter of the overall trend in decentralization. Endogeneity concerns are addressed in an IV framework using a counterfactual transport network as an instrument.

Introduction

Accessibility is one of the most striking determinants of urban structure. Both early rent theory (Alonso, 1964, Mills, 1969, Muth, 1969), as well as more recent polycentric models, acknowledge the trade-off between transport cost, and access to firms, employment opportunities, employees, and customers, from which land values emerge. Some of the largest shocks to urban accessibility patterns and, hence, potentially to land values occurred during the second half of the 19th century in many cities in the industrialized world. Emerging railway networks defined completely new patterns of accessibility and travel behavior. In this realm we use a new archival data set on commercial land values to investigate the case of Berlin (1890–1936), one of the most important and fastest growing cities in continental Europe, during the period of industrialization. We exploit a period where the rapid transit network, consisting of numerous metro-rail and suburban railway lines, was largely established as the backbone of urban transport and considerably reduced travel time and cost. As shown in our analyses, Berlin represented a typical monocentric city that can be investigated within the framework of traditional rent theory. Based on standard comparative statics of the monocentric model we hypothesize that a) there was significant decentralization over the time studied in terms of distance to the CBD, b) there was less decentralization in terms of travel time to the CBD and c) decentralization in terms of distance can partially be explained by travel time reductions.

We highlight that we contribute to a still relatively limited body of evidence. There is a long tradition of empirically testing the theoretical predictions of the monocentric economy. However, due to severe data limitations, the majority of studies have been conducted on the basis of population density measures (e.g. Clark, 1951, McDonald, 1989, McMillen, 1994) - although, arguably, the estimation of a bid-rent curve on the basis of land values represents the most direct way of testing the monocentric city model. Effectively, there is still very little evidence available using historical land values outside the U.S. or, more precisely, outside Chicago.2 The rich body of excellent studies using the historical land value data for Chicago is owed to two unique data sources; “100 Years of Land Values”, by Hoyt (1933), who provides land values on square mile tracks from 1836 to 1928, and “Olcott's Land Values Blue Book of Chicago”, presenting annual estimates for land values at block level since the early 1900s. Aside from these, hardly any comparable data sources could be found in the literature.3 The new archival data for Berlin, Germany, used in this study, is comparable in quality to that for Chicago and is even somewhat more detailed. It is taken from precise city maps at the level of individual plots (1890–1910) and extensive street indices (1928–1936). After digitizing and geo-processing these sources for use in empirical analyses, we are able to add evidence to the evolution of city structure for one of the most dynamic cities in continental Europe during the period of industrialization. In particular we provide evidence for a city with a much longer history and therefore a more established city structure than the few cities for which evidence is currently available. Our case is also unique in the sense that we investigate a capital city with numerous administrative entities that increase the relative importance of the CBD. Most importantly, our analyses make use of digitally-processed network plans, tracking the evolution of the urban rapid transit system, whose level of detail is unique in an historical context. Additionally, we construct a counterfactual least distance network based on exogenous planning objectives, which we use as an instrument in a 2SLS IV strategy in order to address possible endogeneity. The data set is presented in the next section followed by the empirical analyses. First, we review standard land gradient estimates for our study area in the realm of international evidence (Section 3). Second, we provide evidence in support of our three hypotheses described above (Section 4). The last section concludes.

Section snippets

Industrializing Berlin 1890–1936

At the study's starting point, Prussia in general, and especially Berlin, had entered the second phase of industrialization. As is typical for industrializing regions, the revolutionary changes in production technologies generated enormous demand for a larger labor force, drawing peasants and villagers into the fast-growing cities. In addition, Berlin, after 1871, held the status of the capital for both Prussia and the German Reich, which was formed after the victorious French–Prussian war.

Land gradient evolution in international comparison

This section provides estimates of historical land gradients in Berlin, which are discussed against the background of international evidence. The point of departure for our empirical analysis is the standard monocentric city model in which firms bid higher prices for land closer to an exogenous CBD due to lower transport costs and/or travel time savings (Alonso, 1964, Mills, 1969, Muth, 1969). Evidence suggesting that land values (LV) may be well-described by an exponential function of distance

The impact of accessibility

In this section we turn our attention to some simple hypotheses derived from the monocentric city model regarding the adjustment of land values to travel cost reductions: a) A reduction in transport cost relatively increases the bids for parcels of land in the urban periphery at the expense of central locations, which is mirrored in a flattening land gradient in terms of distance. b) Assuming that perceived transport cost are largely determined by the opportunity cost of travel time, the

Conclusion

In this study we have provided evidence for the positive impact of accessibility on the productivity of commercial land, using the evolution of the rapid transport network in Berlin, from 1890 to 1936, as a case study. We have introduced a new archival land-value data-set, which is unique in terms of detail for European, historically grown, and capital cities. We have shown that a) over the study period the city experienced a considerable degree of decentralization, which was a typical

Acknowledgments

We thank Stephen Redding, Daniel Sturm and Nikolaus Wolf for excellent research environments in Berlin and London. We acknowledge the hospitality of the CEP at the London School of Economics, where a considerable part of this research was conducted. We thank Edward Glaeser, Daniel McMillen, Volker Nitsch, Stuart Rosenthal, Daniel Sturm, and two anonymous referees for very helpful comments. We furthermore thank the participants of the KOF research seminar (ETH Zurich), the conference of the

References (45)

  • D.P. McMillen et al.

    Selection bias and land development in the monocentric city model

    Journal of Urban Economics

    (1992)
  • P. Abelson

    House and Land Prices in Sydney from 1931 to 1989

    Urban Studies

    (1997)
  • G.M. Ahlfeldt

    If Alonso was right: modeling accessibility and explaining the residential land gradient

    Journal of Regional Science

    (2010)
  • W. Alonso

    Location and land use: toward a general theory of land rent

    (1964)
  • J. Atack et al.

    Location, location, location! The price gradient for vacant urban land: New York, 1835 to 1900

    Journal of Real Estate Finance & Economics

    (1998)
  • Aust, B. (Cartographer). (1986). Die städtebauliche Entwicklung...
  • R.S. Bednarz

    The Effect of Air Pollution on Property Value in Chicago

    (1975)
  • K. Beier

    S-Bahn Berlin der neue Triebzug ET 480

    (1990)
  • P. Bley

    Berliner S-Bahn

    (2003)
  • E. Bohm

    Kurfürstendamm. Entstehung und erste Entwicklung

    (1980)
  • BorchertF. et al.

    Berlin und seine S-Bahn

    (1987)
  • C. Clark

    Urban population densities

    Journal of the Royal Statistical Society. Series A

    (1951)
  • Cited by (0)

    1

    Tel.: + 49 6151 166506; fax: + 49 6151 165652.

    View full text