Valuing Rail Access Using Transport Innovations

Summary

"Valuing Rail Access Using Transport Innovations" (Journal of Urban Economics 57(1), 2005) by Stephen Gibbons (London School of Economics) and Stephen Machin (then Professor of Economics at University College London, also a research associate of LSE's Centre for Economic Performance, where the paper first circulated as CEP Discussion Paper No. 0611) is among the most widely cited causal studies of transport capitalization. The paper's abstract frames its contribution directly: it sets out "to ascertain how much consumers value rail access" by studying "the effects on house prices of a transport innovation that altered the distance to the nearest station for some households, but left others unaffected"; "[u]sing the innovation to implement a quasi-experimental approach studying house price changes in affected versus unaffected areas allows us to avoid the biases inherent in cross-sectional valuation work." Where earlier work simply correlated house prices with distance to stations — leaving open the objection that stations are built in already-valuable places — Gibbons and Machin exploit the timing of two late-1990s London rail expansions, the Jubilee Line Extension and the Docklands Light Railway extension, as quasi-experiments: they compare price changes in areas that gained station access against otherwise similar areas whose access did not change.

Findings

The study finds that properties in areas gaining rail access appreciated relative to comparison areas, with the effect confined to homes within about 2 km of a station (the paper's walking-distance cutoff) and increasing with the size of the station-distance reduction. The data are Nationwide Building Society housing transactions for 1997–2001, aggregated to postcode units and split before (1997–1999) and after (2000–2001) the 1999 Jubilee Line Extension and DLR extension openings; the treatment group is postcode units whose distance to the nearest station fell and ended up within 2 km. In the headline difference-in-difference comparison, log prices rose by 0.45 in treated areas against 0.37 in control areas — in the authors' words, "showing house price growth to be roughly 9.3 percentage points (= [exp(.089) – 1]×100) higher in areas affected by the transport innovation" (t-ratio 3.46). This is the source of the ~9% headline figure commonly quoted for the study. In regression form, "prices rose in areas affected by the Jubilee Line and Docklands Light Railway by around 5.5% for each kilometre reduction in station distance close to the new stations (within 2km)," falling to "around 2.1% of property prices" per kilometre once full property and neighbourhood controls and differential spatial time trends are included, and to a most-conservative 1.5% per kilometre in the South East London subsample; "[t]here is no statistically significant impact from distance reductions to postcodes that are beyond 2km from the new stations." The paper also runs conventional cross-sectional specifications on the same data as a benchmark — there "[p]rices fall at the rate of 8.9% per kilometre up to 2km," estimates "substantially higher than those from the innovation models" — direct evidence that correlational designs overstate the rail premium, the bias the quasi-experiment exists to avoid. The valuations are sizable relative to other local amenities studied in the same hedonic literature: the most conservative estimates "indicate that prices go up by 1.5% for a one-quarter standard deviation reduction in station distance," versus "under 0.75%" for a one-quarter standard deviation improvement in local primary school performance. The authors also translate the estimated price effects into an implied value of travel time savings (roughly £1.60 to £6 per hour across specifications, bracketing the £4/hour then recommended to the UK Department for Transport), a step relevant to appraising how much of a transit project's cost could in principle be recovered from the value it creates nearby.

Relation to the Georgist Case

This is direct, credibly identified evidence for the mechanism at the heart of land value capture: public transport investment creates location value that accrues to nearby landowners who did nothing to earn it — the unearned increment in its cleanest modern form. It supports the outcome that public investment capitalizes into nearby land values, and implies that a land value tax or capture instrument could recover part of the cost of transit from the value it creates, as the Henry George Theorem suggests in the limit.

Nuances and Limits

• The estimates are for property values (land + structure bundled); the land-value share of the uplift is not separately identified — a general limitation of capitalization studies.
• Results are London-specific; the Mohammad et al. meta-analysis shows uplift estimates vary widely across cities, rail types, and distances.
• Capitalization measures local value creation; it does not by itself establish the net social return of the investment (some uplift may be displaced from elsewhere).
• Magnitudes on this page are verified against the authors' full working-paper text (CEP Discussion Paper No. 0611, January 2004), not the published journal tables, which remained paywalled at ScienceDirect when this page was reviewed; the published JUE version is the same study, but readers quoting exact point estimates should note the working-paper provenance. Note also that secondary summaries disagree on "the" effect size largely because the paper reports several magnitudes in different units — a ~9.3 percentage-point relative price rise for treated areas overall, versus per-kilometre gradients of 1.5% to 5.5% depending on controls and sample — so any single quoted number should specify which estimate it is.

Bears On

• Outcome: Public investment capitalizes into nearby land values — one of the cleanest quasi-experimental designs in that literature.
• Concept: Land Value Capture — the financing policy this evidence underwrites; Tax Capitalization — the general mechanism.

See Also

• Mohammad et al. (2013) — the systematic evidence across 100+ estimates
• Oates (1969) — the founding capitalization study
• Henry George Theorem — the theory this evidence approximates
• Narrative: The Unearned Increment — the argument it powers

Sources

1. Stephen Gibbons & Stephen Machin (2005), "Valuing Rail Access Using Transport Innovations," Journal of Urban Economics 57(1): 148–169. DOI · CEP Discussion Paper No. 0611 (January 2004), full text · LSE Research Online copy — the CEP working-paper full text was read directly for this page and is the source of all quotations and point estimates above (the 9.3 percentage-point difference-in-difference, the 1.5–5.5% per-km gradients, the 2 km cutoff, the 8.9%/km cross-sectional benchmark, and the school-quality comparison); the published ScienceDirect version was paywalled this session, so the published tables were not compared line-by-line against the working paper (see Nuances and Limits).