# RECORD 1908d2ee0b0995f38129f51d7446a3626531b8d6
# TITLE: Mind the Anticipation Gap: A Difference-in-Differences Analysis of the Elizabeth Line’s Impact on London Housing Prices
# YEAR: 2026.0  VENUE: Vietnamese Economic Review  DOI: 10.67913/vner/2026.1.1.1
# AUTHORS: T. Nguyen
# PAGES: 15  CANDIDATE TABLE PAGES (1-based): [6, 8, 10, 11]
# ABSTRACT: This paper estimates the causal impact of London’s Elizabeth Line on residential property prices using a Difference-in-Differences framework applied to 1,602,487 transactions from the UK Land Registry over 2000–2024. Treated properties within five distance bands of stations are compared to a baseline beyond 10 km, separately for the construction window (2007–2015) and the post-opening window (2022–2024). During construction, prices within 1 km of stations rose by 4.9% relative to control, with substantially larger gains of 7.7–8.1% in the 2–10 km bands, indicating broad anticipatory capitalisation and spillover effects. Following the May 2022 opening, station-adjacent prices fell by 2.0%, consistent with overshooting during the construction phase, while outer-band prices remained unchanged. Effects are concentrated in zones previously under-served by Transport for London. The findings have implications for transport appraisal, value-capture timing, and equity-oriented infrastructure planning.


===== PAGE 1 (score 9, numeric-paren 0, table-words 0, var-words 9) =====

Vietnamese Economic Review
https://doi.org/10.67913/vner/2026.1.1.1
Mind the Anticipation Gap:
A Difference-in-Differences Analysis
of the Elizabeth Line’s Impact
on London Housing Prices
Tien Phong Nguyen 1
1 King’s Business School, King’s College London
Abstract
This paper estimates the causal impact of London’s Elizabeth Line on residential
property prices using a Difference-in-Differences framework applied to 1,602,487
transactions from the UK Land Registry over 2000–2024.
Treated properties
within five distance bands of stations are compared to a baseline beyond 10 km,
separately for the construction window (2007–2015) and the post-opening win-
dow (2022–2024). During construction, prices within 1 km of stations rose by
4.9% relative to control, with substantially larger gains of 7.7–8.1% in the 2–
10 km bands, indicating broad anticipatory capitalisation and spillover effects.
Following the May 2022 opening, station-adjacent prices fell by 2.0%, consis-
tent with overshooting during the construction phase, while outer-band prices
remained unchanged. Effects are concentrated in zones previously under-served
by Transport for London. The findings have implications for transport appraisal,
value-capture timing, and equity-oriented infrastructure planning.
Keywords: Elizabeth Line; Transport infrastructure; House prices.
JEL Classification: R41, R31, R42, H54
1
Introduction
Urban transport infrastructure shapes residential property values primarily
through accessibility and commuting-time effects (Gibbons & Machin 2005).
Empirical work consistently documents premiums for properties near transit,
with effect sizes mediated by network design, anticipation behaviour, and pre-
existing accessibility (Mohammad et al. 2017; Ahlfeldt 2013; Billings 2011). The
Crossrail project—rebranded the Elizabeth Line on opening—offers an unusually
well-defined natural experiment: an £18 billion east–west rail link spanning 118
km from Reading to Shenfield, formally approved in 2007, under construction
Corresponding author(s): Tien Phong Nguyen, King’s Business School, King’s College London,
tien_phong.nguyen@kcl.ac.uk
1


===== PAGE 6 (score 19, numeric-paren 0, table-words 6, var-words 4) =====

6
Mind the Anticipation Gap
is the single strongest cross-sectional predictor of log price (R2 = 0.89), substan-
tially exceeding the explanatory power of macroeconomic indicators. Inflation,
interest rates, and GDP growth show statistically significant but quantitatively
small bivariate relationships with prices, consistent with the dominance of lo-
calised factors in urban housing markets (Ahlfeldt 2013).
Table 1: Descriptive statistics by distance band, 2000–2024
Distance band
Observations
Share
(%)
Mean
log
price
SD log
price
Mean
distance
(km)
School
(km)
TfL
station
(km)
Within 1 km
127,028
7.93
12.71
0.88
0.64
1.48
2.54
1–2 km
194,381
12.13
12.67
0.80
1.51
2.43
3.87
2–5 km
480,837
30.01
12.76
0.81
3.38
1.71
2.97
5–10 km
433,849
27.07
12.77
0.73
7.22
0.43
1.80
More than 10 km
366,392
22.86
12.64
0.66
13.51
0.37
2.59
Total
1,602,487
100.00
12.72
0.76
6.29
1.13
2.64
Notes: Sample includes residential transactions in Greater London and outer boroughs contain-
ing Elizabeth Line stations, 2000–2024. Distance refers to Euclidean distance to the nearest
Elizabeth Line station; School and TfL station refer to nearest non-Elizabeth amenity. Log
prices are natural logarithms of nominal sale prices in pounds sterling.
4
Empirical Strategy
4.1
Difference-in-Differences Specification
The baseline DiD model is given by equation 1:
ln(Price)it = α + β(Treatedi × Postt) + γXit + δt + ηi + εit
(1)
where Treatedi is a binary indicator for properties within 2 km of an Elizabeth
Line station, Postt is an indicator for the post-treatment period (either post-2007
construction or post-2022 opening), Xit is a vector of property and macroeco-
nomic controls, δt are year fixed effects absorbing common shocks, and ηi are
property fixed effects absorbing unobserved time-invariant heterogeneity. The
coefficient of interest β captures the causal impact of proximity to the Elizabeth
Line, expressed in approximate percentage terms.
The richer specification replaces the binary treatment with distance-band indi-
cators:
ln(Price)it = α +
4
∑
d=1
βd(DistanceBanddi × Postt) + γXit + δt + ηi + εit
(2)
where DistanceBanddi takes values for the within 1 km, 1–2 km, 2–5 km, and 5–
10 km bands relative to the omitted reference of more than 10 km. The vector βd


===== PAGE 7 (score 16, numeric-paren 1, table-words 3, var-words 9) =====

Tien Phong Nguyen
7
traces the spatial profile of the treatment effect. Equations 1 and 2 are estimated
separately for two windows. The construction window covers 2000–2015 with
Post = 1 from 2007. The opening window covers 2016–2024 with Post = 1 from
2022. This split isolates anticipatory capitalisation, when the line was authorised
but not running, from post-opening adjustment, when accessibility gains were
physically realised.
4.2
Parallel Trends
Identification requires that, absent treatment, treated and control properties
would have followed parallel price paths.
The assumption is assessed with
placebo regressions estimating year-by-year coefficients in the pre-treatment
windows (2000–2006 for the construction phase; 2016–2021 for the opening
phase). Pre-construction estimates show closely aligned price trajectories across
distance bands.
For example, in 2001 the log price coefficient for properties
within 1 km of future stations is 0.022, virtually identical to the 1–2 km band
(0.025), and the alignment holds through 2006.
In the opening pre-period
(2016–2021), modest divergences emerge from 2020 onward—properties within
1 km record a small negative coefficient of -0.030 in 2021 (p = 0.032), reflecting
anticipatory adjustment as the operational launch approached.
This pattern
is consistent with prior literature documenting anticipatory effects ahead of
major transport openings (Gibbons & Machin 2005; Agostini & Palmucci 2008)
and reflects rational expectations rather than a violation of parallel trends.
Robustness checks below confirm that the main DiD estimates are not driven by
these end-of-window deviations.
4.3
Event Study
To trace the temporal profile of treatment effects, I estimate the event study in
equation 3:
ln(Price)it = α +
−2
∑
T=−q
βT(Treatedi × DT
t ) +
m
∑
T=0
βT(Treatedi × DT
t )
+ γXit + δt + ηi + εit
(3)
where DT
t indicates calendar year T, and the year preceding treatment (T =
−1) is omitted as the baseline. The βT coefficients capture year-specific devi-
ations between treated and control properties. The event-study design enables
tests for pre-trends, identifies the timing of capitalisation, and characterises post-
treatment dynamics.
4.4
Robustness
Several robustness exercises are conducted. First, property-type indicators (de-
tached, semi-detached, terraced, flat), estate type, and amenity distances are


===== PAGE 8 (score 27, numeric-paren 11, table-words 5, var-words 4) =====

8
Mind the Anticipation Gap
added to the baseline specification. Second, borough socioeconomic controls are
added for the 2017–2024 sub-window. Third, heterogeneity is tested by inter-
acting the treatment with an indicator for pre-existing TfL access, defined as a
non-Elizabeth TfL station within 1 km. Fourth, for the post-2017 sub-window
I compare fixed-effects (FE) and random-effects (RE) specifications. The Haus-
man test is inconclusive because of a non-positive-definite variance-covariance
matrix—a common outcome when within-property variation is limited (Moham-
mad et al. 2017). The Breusch–Pagan LM test (p < 0.001) and a moderate rho
of 0.36 motivate FE as the primary specification while reporting RE estimates
as a secondary check, consistent with prior DiD work (Gibbons & Machin 2005;
Billings 2011). Standard errors are clustered at the property level throughout.
5
Results
5.1
Construction Phase, 2007–2015
Table 2: DiD estimates: construction phase, 2007–2015
Binary (1)
Distance bands (2)
Treated × Construction
−0.003
(0.002)
Within 1 km × Construction
0.049
(0.003)
1–2 km × Construction
0.053
(0.002)
2–5 km × Construction
0.081
(0.002)
5–10 km × Construction
0.077
(0.002)
Inflation rate
0.320
−0.006
(0.017)
(0.001)
Interest rate
−6.055
0.060
(0.243)
(0.005)
GDP growth rate
0.710
0.032
(0.025)
(0.003)
Property and year fixed effects
Yes
Yes
Observations
541,796
541,796
Notes: Dependent variable is the natural logarithm of nominal sale
price. Standard errors clustered at the property level shown in paren-
theses. Reference category in column (2) is properties more than 10
km from an Elizabeth Line station. Coefficients on “Yes”-valued rows
indicate inclusion of the corresponding fixed effects; standard errors
omitted for these rows.
Table 2 reports DiD estimates for the construction window. The binary specifi-
cation in column (1) yields a small and only marginally significant coefficient of
−0.003 (p = 0.063), suggesting little aggregate price movement when spatial


===== PAGE 9 (score 11, numeric-paren 0, table-words 3, var-words 5) =====

Tien Phong Nguyen
9
heterogeneity is collapsed. Column (2) replaces the binary treatment with the
four distance-band interactions and tells a substantively different story. All four
coefficients are positive and significant at the 1% level. Properties within 1 km
of an Elizabeth Line station appreciated by approximately 4.9% relative to the
more-than-10-km reference group; the 1–2 km band gained 5.3%. The largest
effects, however, are in the mid-range bands: 8.1% in the 2–5 km band and 7.7%
in the 5–10 km band. Macroeconomic controls enter with the expected signs.
The non-monotonic distance gradient is economically interpretable. Mid-range
zones may have benefited from a combination of expected accessibility gains and
limited exposure to construction disamenities such as traffic, noise, and ground
works concentrated near station sites. Properties in the immediate vicinity were
more exposed to these short-term costs, partially offsetting the anticipated long-
term benefits. The pattern echoes findings from Agostini & Palmucci (2008) and
Dubé et al. (2013), who document larger anticipation effects in zones marginally
removed from new infrastructure. The construction-phase event study traces this
pattern over time: year-specific coefficients for properties within 2 km of stations
turn positive shortly after the 2007 announcement, rise through the early 2010s,
and reach their peak in 2014–2016, foreshadowing the correction documented
below. Figure 2 presents the event-study coefficients for the full 2000–2024 hori-
zon.
Figure 2: Event-study estimates for properties within 2 km of an Elizabeth Line station, 2000–
2024. The y-axis is the year-specific log-price coefficient relative to the reference year (2006).
5.2
Post-Opening Phase, 2022–2024
Table 3 reports DiD estimates for the opening window. The binary specification
in column (1) gives a coefficient of −0.014 (p = 0.001), indicating that prices
within 2 km of a station are 1.4% lower than control prices following the May
2022 launch. Column (2) disaggregates by distance band: properties within 1 km


===== PAGE 10 (score 26, numeric-paren 11, table-words 3, var-words 7) =====

10
Mind the Anticipation Gap
decline by 2.0% relative to the reference (p = 0.005); the 1–2 km band declines
by 1.2% (p = 0.024). Coefficients for the 2–5 km and 5–10 km bands are statis-
tically indistinguishable from zero. The post-opening profile is the spatial mirror
of the construction-phase result: the correction is sharpest in the bands that expe-
rienced the largest pre-opening anticipation, particularly the inner bands where
construction-disamenity offsets had limited the apparent uplift.
Table 3: DiD estimates: post-opening phase, 2022–2024
Binary (1)
Distance bands (2)
Treated × Opening
−0.014
(0.004)
Within 1 km × Opening
−0.020
(0.007)
1–2 km × Opening
−0.012
(0.005)
2–5 km × Opening
−0.001
(0.004)
5–10 km × Opening
−0.003
(0.004)
Inflation rate
−0.002
−0.002
(0.001)
(0.001)
Interest rate
0.019
0.019
(0.002)
(0.002)
GDP growth rate
0.014
0.014
(0.001)
(0.001)
Property and year fixed effects
Yes
Yes
Observations
170,093
170,093
Notes: Dependent variable is the natural logarithm of nominal sale
price. Standard errors clustered at the property level shown in paren-
theses. Reference category in column (2) is properties more than 10
km from an Elizabeth Line station. The Opening indicator equals one
from 2022 onwards.
The opening-phase event study confirms this interpretation.
The treated–
control gap, having peaked in 2014–2016, narrows steadily from 2017 onward
and turns modestly negative for inner-band properties from 2022. By 2023–2024,
treated coefficients are close to zero, consistent with the market settling at or
slightly below the level implied by realised accessibility gains. The pattern is
consistent with rational expectations and information efficiency:
anticipated
benefits were priced in well before service commenced, leaving little scope
for further gains and some scope for correction of mid-2010s overshooting.
Comparable post-opening reversion has been documented around Dubai Metro
openings (Mohammad et al. 2017) and in commuter rail studies (Dubé et al.
2013).


===== PAGE 11 (score 19, numeric-paren 6, table-words 4, var-words 4) =====

Tien Phong Nguyen
11
5.3
Heterogeneity by Pre-existing Transport Access
A central question for value-capture and equity policy is whether the Elizabeth
Line’s price effects vary with pre-existing accessibility. Equation 4 tests this by
adding interactions between the treatment and an indicator TfLi for whether a
property is within 1 km of a non-Elizabeth TfL station:
ln(Price)it = α + β1(Treatedi × Postt) + β2(TfLi × Postt)
+ β3(Treatedi × TfLi × Postt) + γXit + δt + ηi + εit
(4)
Table 4 reports the construction-phase and opening-phase estimates from equa-
tion 4. For properties near new Elizabeth Line stations but without prior TfL cov-
erage, the construction-period effect is negative (−0.038, p < 0.001), indicating
that early construction disruption was costly for these previously under-served
locations. The triple interaction with prior TfL access is positive and significant
(0.021, p < 0.001), indicating that pre-existing TfL connectivity partly insulated
nearby properties from construction-period disamenities. In the opening window,
the triple interaction is positive but not statistically significant (0.015, p = 0.25),
indicating that by 2022 the marginal value of additional TfL access at Elizabeth
Line stations had essentially dissipated.
Table 4: Heterogeneity by pre-existing TfL access
Construction (1)
Opening (2)
Treated × Post
−0.038
−0.021
(0.004)
(0.010)
TfL × Post
0.105
−0.034
(0.001)
(0.003)
Treated × TfL × Post
0.021
0.015
(0.006)
(0.013)
Property and year fixed effects
Yes
Yes
Macro and amenity controls
Yes
Yes
Observations
541,796
170,093
Notes: TfL is an indicator equal to one for properties within 1 km
of a non-Elizabeth TfL station. Treated indicates properties within
2 km of an Elizabeth Line station. Standard errors clustered at the
property level shown in parentheses. Column (1) covers 2007–2015;
column (2) covers 2022–2024.
Event-study estimates by sub-group reinforce this. Properties within 1 km of
an Elizabeth Line station that were already well served by TfL show steady appre-
ciation through 2004–2010 followed by gradual decline; the Elizabeth Line adds
little marginal value, consistent with a saturation hypothesis. Properties within 1
km that were not previously connected to TfL show delayed but stronger growth,
peaking around 2014 through 2018, before stabilising post-opening. The het-
erogeneity is consistent with diminishing marginal returns to accessibility and
with international evidence that infrastructure adds the most value in previously
under-served areas (Ahlfeldt & Wendland 2011; Koster et al. 2014).


===== PAGE 12 (score 14, numeric-paren 0, table-words 1, var-words 12) =====

12
Mind the Anticipation Gap
5.4
Macroeconomic and Property-Level Controls
Macroeconomic controls perform consistently with theory but should be inter-
preted cautiously because year fixed effects absorb most national-level variation.
Inflation and interest rates are negatively associated with prices in both windows;
GDP growth is positively associated. Magnitudes are larger in the construction
window, reflecting wider macroeconomic variation across 2000–2015 (including
the 2008–2009 recession). In the post-2017 window, broader shocks—the Brexit
referendum, the COVID-19 pandemic, and the 2022 monetary tightening—appear
to have suppressed further appreciation in zones where anticipatory capitalisa-
tion had already occurred, consistent with Vergara-Perucich (2024) and Hilber &
Vermeulen (2016).
Property-type controls cannot be identified in the FE specification because prop-
erty type is time-invariant. In the RE specification estimated for the 2017–2024
sub-window for which all controls are observed, flats show the largest negative
coefficient relative to the omitted base (−0.75, p = 0.003), consistent with flats
in already-dense central locations being most exposed to overshooting and sub-
sequent correction. Detached homes show −0.63 (p = 0.003); terraced homes
show −0.40 (p = 0.002). These results should be read as conditional compar-
isons rather than causal effects, given the omitted within-property variation in
FE.
6
Discussion
The estimates support three substantive conclusions.
First, the Elizabeth
Line generated economically meaningful capitalisation effects, but these were
concentrated in the construction phase rather than the operational phase.
Forward-looking buyers priced expected accessibility gains long before service
commenced, consistent with the rational-expectations capitalisation framework
(Agostini & Palmucci 2008; Ahlfeldt 2013).
Second, the spatial profile of
anticipation is non-monotonic:
mid-range zones (2–10 km) experienced the
largest construction-period gains, plausibly because they combined expected
accessibility benefits with limited exposure to construction disamenities. Third,
the post-opening correction is concentrated in inner bands, suggesting that an-
ticipation in those zones overshot the realised improvement in net accessibility
once disamenities, congestion, and competing routes are taken into account. The
overshooting–correction pattern echoes findings around Dubai Metro openings
(Mohammad et al. 2017) and other rail studies (Dubé et al. 2013).
The heterogeneity by pre-existing TfL access has direct implications for
transport-led equity and value-capture policy.
The Elizabeth Line generated
larger and more durable price responses in zones not previously served by TfL,
consistent with the diminishing marginal returns to accessibility predicted by
bid-rent theory and confirmed in international evidence (Ahlfeldt & Wendland
2011; Koster et al. 2014). This argues for prioritising future transit investment
in under-served catchments where capitalisation effects—and the social welfare
gains they signal—are largest. Three caveats are important. First, distance-based