pith. sign in

arxiv: 2408.01562 · v2 · submitted 2024-08-02 · 💻 cs.CY

Welfare, sustainability, and equity evaluation of the New York City Interborough Express using spatially heterogeneous mode choice models

Hai Yang , Hongying Wu , Lauren Whang , Xiyuan Ren , Joseph Y. J. Chow This is my paper

Pith reviewed 2026-05-23 21:53 UTC · model grok-4.3

classification 💻 cs.CY
keywords light rail evaluationmode choice modeltransit ridership forecasturban GHG reductionequity in transitNew York City mobilityconsumer surplus
0
0 comments X

The pith

The proposed Interborough Express light rail would save 28.1 minutes per trip on average and draw more than 272 thousand daily riders.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper applies a block-group-level mode choice model to synthetic trip data to forecast effects of the new IBX light rail linking Queens and Brooklyn. It projects citywide time savings, ridership far above the official estimate, mode shifts from cars, daily GHG cuts of 30.63 metric tons, and $0.89 per trip consumer surplus, with 21.4 percent of riders from low-income households. A sympathetic reader would care because these numbers indicate the project could expand transit access and cut emissions if the forecasts hold.

Core claim

Using a spatially heterogeneous mode choice model calibrated on open synthetic trip data, the paper concludes that the IBX would generate over 272 thousand daily riders (81 percent above the official proposal), attract more than 40 thousand new transit trips including over 16 thousand from private vehicles, reduce GHG emissions by 30.63 metric tons per day, deliver $0.89 USD consumer surplus per trip, and serve 21.4 percent low-income riders while failing to meaningfully shrink the share of travelers with very low surpluses.

What carries the argument

Block-group-level spatially heterogeneous mode choice model calibrated on open-access synthetic trip agenda dataset, used to predict traveler responses to the new service.

If this is right

  • IBX would save 28.1 minutes to potential riders across the city and 29.7 minutes for those near the line.
  • More than 185 thousand daily riders would start or end along the IBX corridor.
  • Over 58 thousand daily riders (21.4 percent) would come from low-income households.
  • The addition of IBX would shift more than 16 thousand daily trips from private vehicles.
  • Consumer surplus benefits reach $0.89 USD per trip but do not significantly reduce the share of travelers with surpluses below 10 percent of the average.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the ridership and mode-shift numbers hold, the IBX could serve as a template for evaluating other cross-borough transit projects using similar spatially detailed models.
  • The limited equity impact on the lowest-surplus travelers implies that complementary fare or service policies may be needed to broaden benefits.
  • Citywide GHG savings of 30 tons per day could scale with additional lines if the underlying mode choice parameters generalize.

Load-bearing premise

The block-group-level spatially heterogeneous mode choice model calibrated on the synthetic trip agenda dataset produces accurate forecasts of how travelers would actually respond to the new IBX service.

What would settle it

Comparison of the model's projected daily ridership, mode shares, and time savings against observed counts and survey data collected after the IBX opens and operates.

Figures

Figures reproduced from arXiv: 2408.01562 by Hai Yang, Hongying Wu, Joseph Y. J. Chow, Lauren Whang, Xiyuan Ren.

Figure 1
Figure 1. Figure 1: The Transit Connectivity Index (TCI) in (a) Manhattan, (b) Brooklyn, and (c) Queens (source: CNT, 2023) [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Proposed route design of the IBX With many potential benefits expected from the IBX, there are limited studies quantifying these benefits comprehensively. There is limited survey data collected specifically for this route, so a regional planning model is needed. However, existing models generally do not capture spatial heterogeneity in taste variation. In other words, the value of time for one income level… view at source ↗
Figure 3
Figure 3. Figure 3: General logics of the MXL, IPL, and GLAM logit models [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 1
Figure 1. Figure 1: With the introduction of IBX, higher transit [PITH_FULL_IMAGE:figures/full_fig_p013_1.png] view at source ↗
read the original abstract

The Metropolitan Transit Authority (MTA) proposed building a new light rail route called the Interborough Express (IBX) to provide a direct, fast transit linkage between Queens and Brooklyn. An open-access synthetic citywide trip agenda dataset and a block-group-level mode choice model are used to assess the potential impact IBX could bring to New York City (NYC). IBX could save 28.1 minutes to potential riders across the city. For travelers either going to or departing from areas close to IBX, the average time saving is projected to be 29.7 minutes. IBX is projected to have more than 272 thousand daily ridership after its completion (81% higher than reported in the official IBX proposal). Among those riders, more than 58 thousand people (21.4%) would come from low-income households while 185 thousand people (68.2%) would start or end along the IBX corridor. The addition of IBX would attract more than 40 thousand additional daily trips to transit mode, among which more than 16 thousand would be switched from using private vehicles, reducing potential greenhouse gas (GHG) emissions by 30.63 metric tons per day. IBX can also bring significant consumer surplus benefits to the communities, which are estimated to be $0.89 USD per trip. However, the service does not appear to significantly reduce the proportion of travelers whose consumer surpluses fall below 10% of the population average (already quite low).

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper uses an open-access synthetic citywide trip agenda dataset together with a block-group-level spatially heterogeneous mode choice model to project the impacts of the proposed Interborough Express (IBX) light-rail line in New York City. It reports average time savings of 28.1 minutes (29.7 minutes near the corridor), daily ridership above 272 000 (81 % above the official estimate), more than 40 000 new transit trips (including >16 000 diverted from private vehicles), a daily GHG reduction of 30.63 t, and a consumer-surplus benefit of $0.89 per trip, with 21.4 % of riders from low-income households.

Significance. If the underlying mode-choice forecasts are shown to be reliable, the study supplies a spatially resolved, equity-aware evaluation of a major transit investment that could usefully inform MTA planning and similar corridor analyses elsewhere.

major comments (3)
  1. [Methods] Methods section on model calibration: the block-group-level logit is fitted to the synthetic trip agenda dataset, yet no hold-out validation, comparison against observed MTA ridership counts, or out-of-sample mode-share checks on existing corridors are reported. Because every headline projection (ridership, mode shifts, GHG, surplus) is generated by applying this fitted model to the same synthetic data, the absence of external validation directly undermines the quantitative claims.
  2. [Results / Abstract] Results and abstract: the 81 % higher ridership figure relative to the official proposal is stated without any reconciliation of modeling assumptions, trip-generation rates, or network representations that differ between the two studies; this discrepancy is load-bearing for the central policy-relevant claim.
  3. [Results] No sensitivity or uncertainty analysis is presented for the synthetic dataset's trip-generation or income-group assumptions, even though systematic bias in those inputs would scale all reported ridership, mode-shift, and emission figures proportionally.
minor comments (2)
  1. Figure captions and axis labels should explicitly note which outputs are model predictions versus observed data.
  2. [Abstract] The abstract states 'more than 272 thousand daily ridership' and similar rounded figures; the corresponding tables or text should supply the exact modeled values and any rounding conventions used.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their constructive comments, which highlight important areas for strengthening the manuscript. We agree that additional validation steps and sensitivity analyses will improve the robustness of the projections. Below we respond point-by-point to the major comments and outline planned revisions.

read point-by-point responses

  1. Referee: [Methods] Methods section on model calibration: the block-group-level logit is fitted to the synthetic trip agenda dataset, yet no hold-out validation, comparison against observed MTA ridership counts, or out-of-sample mode-share checks on existing corridors are reported. Because every headline projection (ridership, mode shifts, GHG, surplus) is generated by applying this fitted model to the same synthetic data, the absence of external validation directly undermines the quantitative claims.

    Authors: We acknowledge that external validation would strengthen confidence in the projections. Because the input is a synthetic dataset (open-access and generated to match aggregate patterns), direct access to confidential individual MTA trip records for out-of-sample checks is unavailable. We will add internal hold-out validation by splitting the synthetic data, reporting metrics such as prediction accuracy and log-likelihood on the test portion. We will also compare model-derived aggregate mode shares against publicly available NYC transit statistics for existing corridors and include a brief discussion of how the synthetic data was calibrated to aggregate counts. These changes will appear in a revised Methods section. revision: yes

  2. Referee: [Results / Abstract] Results and abstract: the 81 % higher ridership figure relative to the official proposal is stated without any reconciliation of modeling assumptions, trip-generation rates, or network representations that differ between the two studies; this discrepancy is load-bearing for the central policy-relevant claim.

    Authors: We agree that reconciliation is needed to contextualize the difference. In the revised Results and Discussion sections we will explicitly compare modeling choices: our block-group-level spatial heterogeneity and citywide synthetic trip generation versus the official study’s likely more aggregate assumptions and corridor-focused scope. We will note differences in trip-generation rates, network detail, and the inclusion of all potential citywide trips in our framework, while acknowledging that these choices contribute to the higher estimate. revision: yes

  3. Referee: [Results] No sensitivity or uncertainty analysis is presented for the synthetic dataset's trip-generation or income-group assumptions, even though systematic bias in those inputs would scale all reported ridership, mode-shift, and emission figures proportionally.

    Authors: We accept this limitation and will add sensitivity analyses. In the revised Results we will vary trip-generation rates by ±10 % and ±20 % and adjust income-group shares within census margin-of-error bounds, reporting the resulting ranges for daily ridership, vehicle-to-transit shifts, GHG reductions, and consumer surplus. This will demonstrate the degree to which headline figures scale with input assumptions. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper estimates parameters for a block-group-level spatially heterogeneous mode choice model using an open-access synthetic trip agenda dataset representing base conditions, then applies the fitted model to a counterfactual network that includes the proposed IBX service to generate ridership, mode-shift, time-saving, GHG, and consumer-surplus forecasts. This is a conventional discrete-choice forecasting workflow in which the estimation step matches observed or synthetic base behavior and the prediction step evaluates the same functional form on altered attributes; the two steps are not identical by construction. No equations, self-citations, or uniqueness claims are shown that would reduce any headline output to a fitted input or prior author result. The derivation therefore remains self-contained against external benchmarks once the modeling assumptions are granted.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

Evaluation rests on the accuracy of a fitted mode choice model and a synthetic trip dataset whose generation process is not detailed in the abstract; no new physical entities are postulated.

free parameters (1)
  • mode choice model parameters
    Spatially heterogeneous coefficients in the block-group-level mode choice model are calibrated to data and directly determine all projected ridership, time savings, and surplus figures.
axioms (1)
  • domain assumption The open-access synthetic trip agenda dataset is representative of actual NYC travel patterns and preferences.
    Used as the sole input for all simulations and impact calculations.

pith-pipeline@v0.9.0 · 5822 in / 1345 out tokens · 26888 ms · 2026-05-23T21:53:22.940364+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

4 extracted references · 4 canonical work pages · 1 internal anchor

  1. [1]

    American Public Transportation Association. (2022). Public transportation fact book. American Public Transportation Association. Bar-Gera, H., Konduri, K. C., Sana, B., Ye, X., & Pendyala, R. M. (2009). Estimating survey weights with multiple constraints using entropy optimization methods (No. 09-1354). Benjamin, J., Kurauchi, S., Morikawa, T., Polydoropo...

    work page 2022

  2. [2]

    McFadden, D., & Train, K. (2000). Mixed MNL models for discrete response. Journal of applied Econometrics, 15(5), 447-470. McNally, M. G. (2007). The four-step model. In Handbook of transport modelling (Vol. 1, pp. 35-53). Emerald Group Publishing Limited. MTA mobility survey (2019). MTA New York City travel surveys https://new.mta.info/transparency/surve...

    work page 2000

  3. [3]

    https://new.mta.info/agency/new-york-city- transit/subway-bus-ridership-2022 [Accessed on 02/10/24] MTA. (2023c). The Interborough Express, planning & Environmental Linkages Study. https://new.mta.info/document/114891 [Accessed on 02/02/24] MTA. (2024). Interborough Express. https://new.mta.info/project/interborough-express [Accessed on 02/02/24] NYCDOT. ...

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  4. [4]

    Wang, J., Zhang, N., Peng, H., Huang, Y ., & Zhang, Y . (2022). Spatiotemporal heterogeneity analysis of influence factor on urban rail transit station ridership. Journal of Transportation Engineering, Part A: Systems, 148(2), 04021115. Wu, J., He, D., Jin, Z., Li, X., Li, Q., & Xiang, W. (2024). Learning spatial–temporal pairwise and high- order relation...

    work page doi:10.5281/zenodo.8231009 2022