{"id":"97c798c0-247a-4c64-bce0-bbed3fe93c88","arxiv_id":"2505.16946","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Applying name- and location-based race imputation to New York property records, the paper estimates that White ownership share exceeds White population share in most tracts, with the largest gaps in minority-majority neighborhoods.","lead":"This paper applies a machine learning model that guesses a property owner's race from name and location, then compares those guesses with census population data in New York census tracts. It reports that White people own a larger share of properties and property value than their population share, particularly in neighborhoods where most residents are Black or Hispanic, and that corporate ownership widens the gap.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 5-10 percentage-point disparity estimates rest on an unvalidated domain transfer, and the paper's own stress test cannot establish robustness because it applies PPP-derived error rates via formulas that are not a valid confusion-matrix inversion.","rationale":"The reader's weakest-assumption diagnosis is the same one I would make: the gap estimates depend on an imputation model that is not validated on the target population. My pass adds two specifics. First, the validation set is PPP applicants, who are not property owners; the model was trained on FL/NC voters, so even the analogous NY PPP validation is only a partial transport check. Second, Section 3.4 does not perform a correct bias correction: FPR is a conditional probability over true labels, and multiplying predicted shares by (1-FPR) is not equivalent to inverting the confusion matrix. This matters because the stress test is the paper's main defense against the objection that White overrepresentation is a misclassification artifact. The paper's directional conclusion is consistent with established homeownership gaps (Comptroller 2021; Furman Center 2023), so I do not think the central claim should be rejected; but the specific 5-10 pp magnitude and the extreme-disparity tables should be treated as conditional pending validation on NY property owners with known race. No change to the reader's CONDITIONAL verdict is needed.","tokens_in":16174,"tokens_out":7257,"duration_ms":62872,"concrete_test":"Compile a held-out sample of NY property owners with self-reported race (e.g., NY voter registration matched to deed/assessment owner names; HMDA mortgage applicant race matched to parcel owners; or a manually labeled random sample of owner names in the NYS/NYC assessment files). Run the Full and Name-Only models on this sample and compute the tract-level signed error in White owner share, stratifying by majority-Black and majority-Hispanic tracts. If the mean signed error is >2 percentage points, or if applying the confusion-matrix-inverted correction to Table 1 moves the White-owner minus White-population gap outside 5-10 points, the central magnitude claim fails. Independently, replace the Section 3.4 stress adjustment with a true inversion of Tables 7 and 8, and check whether the extreme tract White-owner shares remain above the White population share.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing premise of Section 3.1 / Table 1 is that the imputation model's error structure on New York property owners matches the validation setting. It does not: the model was trained on Florida and North Carolina voter registration files (Appendix A.2) and validated on Paycheck Protection Program loan applicants in New York (Appendix B), not on NY property owners with known race. PPP applicants are business owners, not necessarily property owners, and the two populations differ in age, income, name distributions, and geographic composition. Because the headline estimate is a small gap (72% vs 65% White owner/population share, i.e. 5-10 points), even a modest differential misclassification rate in the target population could account for the entire gap. The paper's stress test (Section 3.4) does not fix this: it applies Table 9 White FPR by multiplying predicted White share by (1-FPR) and minority shares by 1/(1-FNR), which is not a confusion-matrix inversion and is statistically invalid; FPR is a true-label conditional probability, not a precision correction, and base rates are ignored. Therefore the claim that extreme disparities persist 'under conservative assumptions' is not established, and the statewide magnitude remains vulnerable to model transfer bias.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper applies an LSTM-based race/ethnicity imputation model (Full Model with geolocation and XGBoost; Name-Only Model) to New York property parcel data, inferring the race of individual property owners and comparing the predicted owner composition with census tract population composition. The authors report that White individuals are overrepresented among property owners relative to their population share in over 81% of tracts, with a statewide mean gap of 5-10 percentage points, and that in majority-Black and majority-Hispanic tracts the majority group owns far less than its population share while White ownership is disproportionately high. They further examine corporate ownership and urban-suburban-rural differences, and perform a stress test intended to show that extreme disparities persist under error corrections. The paper also presents ground-truth validation of the imputation models on Paycheck Protection Program loan applicant data.","tokens_in":16498,"tokens_out":6460,"duration_ms":45523,"significance":"If the quantitative claims were established, the paper would make a useful contribution by providing tract-level estimates of racial ownership disparities in New York, supplementing household-level homeownership statistics with property-level evidence and connecting to policy debates on absentee ownership and financialization. The paper is commendable for comparing two model variants, reporting confusion-matrix-based error rates, and discussing limitations candidly, including the difference between owner-occupied and investor-owned properties. However, the central quantitative estimates rest on an unvalidated domain transfer and on a stress-test procedure that is not a valid confusion-matrix correction. These issues do not necessarily invalidate the qualitative direction of the findings, which is consistent with prior homeownership statistics, but they preclude accepting the reported magnitudes as established. The corporate-ownership analysis also relies on an interpretive assumption that corporate entities can be treated as White-controlled.","major_comments":[{"comment":"The stress test is not a valid confusion-matrix inversion. The White FPR is P(predicted=White | true≠White); multiplying the predicted White share by (1-FPR) assumes that all predicted-White labels that are wrong are true non-White, and ignores the base rate of White owners, so it does not produce an upper bound on the true White share. Similarly, dividing minority shares by (1-FNR) does not correctly recover true minority counts unless the false negative count is corrected using the appropriate base rates. Consequently, the 'Stress' columns in Table 3 do not demonstrate that the disparities persist under conservative assumptions; a valid correction would require applying the full confusion matrix (or its inverse) to the predicted counts, or estimating positive predictive values in the target population.","section":"Section 3.4, Table 3"},{"comment":"The imputation model is trained on Florida and North Carolina voter registration data and validated on New York PPP loan applicants, but is then applied to New York property owners, a population that differs in age, income, geography, and name distribution. No validation on property owners with known race is provided, and the headline statewide gap is only 5-10 percentage points, so even a modest differential misclassification rate in the target population could account for the entire gap. The paper should either provide target-population validation, derive an upper bound on the disparity that is robust to plausible misclassification, or explicitly downgrade the quantitative claims to illustrative.","section":"Appendix A.2, Appendix B, Section 3.1"},{"comment":"The Full Model uses census tract demographics as an input feature, and the comparison benchmark is the same census tract demographics. This creates a circularity: predicted owner composition is partially regressed toward the resident population distribution, so the model-based gap is not an independent measurement. The paper also applies the Full Model statewide but the Name-Only Model in NYC, confounding regional comparisons with model differences. A sensitivity analysis that re-runs the full-state analysis with tract demographics removed, or that compares both models on the same tracts, is needed to assess the size of these effects.","section":"Appendix A.2, Section 3.1"},{"comment":"The combined 'White + Corp' ownership metric is presented as if corporate ownership is White-controlled, but the paper itself acknowledges that 'we cannot attribute race directly' to corporations. Adding corporate ownership to White individual ownership therefore overstates the White-controlled share and should be labeled as a scenario or bound, not a measured quantity. In addition, the text reporting Albany tract 11 (Section 3.5) says White individuals own 41.9% of properties, but Table 5's columns (White+Corp 94.1%, Corp Only 89.8%) imply White individual ownership of only 4.3%; this internal inconsistency must be resolved.","section":"Section 3.5, Table 5"}],"minor_comments":[{"comment":"The sentence 'We apply error rate adjustments from our PPP ground truth validation (Section 3.2, Table 9)' misreferences the appendix; Table 9 appears in Appendix B, not Section 3.2.","section":"Section 3.4"},{"comment":"The header 'T ract (Muni - T ract ID)' contains a typo; it should be 'Tract (Muni - Tract ID)'.","section":"Table 2"},{"comment":"Figures 1-3 are referenced in the text but do not appear in the manuscript, making the visual claims unverifiable; the figures or a note on their availability should be included.","section":"Figures"},{"comment":"The Urban-Core classification uses an ad-hoc threshold of '>1000 properties per tract' without sensitivity analysis; the choice should be justified or tested against alternative thresholds.","section":"Section 3.6"},{"comment":"The median income figures for majority-White and majority-Hispanic tracts ($90k vs $62k) are not reported in any table or tied to a specific ACS year; the source and year should be documented.","section":"Section 4"},{"comment":"The conclusion contains vague statements about a forthcoming technique for estimating error distributions and about the authors having 'conceived a technique'; these passages are unsupported and should be removed or replaced with specific references.","section":"Section 5"},{"comment":"The text references 'Figure 3' for accuracy by income decile, but no figure is provided in the appendix.","section":"Appendix B"}],"recommendation":"major_revision","confidential_remarks":"The manuscript reads as an applied case study rather than a methodological contribution; if the target venue expects novel methodology, the contribution is modest, though the topic is appropriate for a policy-oriented publication. The authors should be encouraged to release the code and processed data to allow replication of the tract-level estimates, as the current manuscript does not provide sufficient detail to reproduce the pipeline without the prior paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead the NY real estate equity paper. Bottom line: the qualitative conclusion is sound—White overrepresentation in property ownership is real and consistent with every homeownership statistic we have—but the quantitative magnitudes are not yet defensible, and the paper's own stress test doesn't fix that.\n\nWhat's actually new: the tract-level comparison of imputed owner race to resident demographics, the catalog of extreme-disparity tracts, and the White-plus-corporate ownership metric. That's a useful lens for planners. The paper is also refreshingly candid about limitations, releases confusion matrices from a PPP validation, and compares a name-only model to a location-aware one. Credit where due.\n\nThe soft spots are real. The imputation model was trained on Florida and North Carolina voter registration data. PPP loan applicants in New York are not property owners; they differ systematically in age, income, and geography. A modest differential misclassification rate could move the headline 5–10 point gap. The Section 3.4 stress test is not a valid correction: multiplying predicted White share by (1−FPR) and minority shares by 1/(1−FNR) ignores base rates and predictive values. FPR is not precision. So the claim that extreme disparities persist 'under conservative assumptions' overreaches.\n\nThere's also circularity: the Full Model takes census tract demographics as an input feature, and the same demographics are the benchmark. That doesn't invalidate the disparity, but it means the comparison isn't fully independent. The NYC results use the Name-Only model, which the authors acknowledge overestimates White ownership; the direction of that bias is actually conservative for their argument, so it's less damaging.\n\nThe tables have some internal inconsistencies (Table 5's corporate percentages don't always reconcile with the text), but those are fixable.\n\nWho is this for? Urban planners, housing equity researchers, and anyone working on algorithmic fairness in administrative data. It's a good case study for a methods discussion, even if the exact magnitudes shift after validation. I'd send it to peer review with major revisions: require a validation on a sample of NY property owners with known race, or at minimum re-estimate using the bounds the confusion matrix actually implies, and drop the pseudo-correction. The paper is worth engaging.\n\nRecommendation: accept for peer review, conditional on serious methodological revision.","headline":"The qualitative finding is almost certainly right, but the magnitudes rest on an unvalidated domain transfer and a stress test that is not a valid correction.","tokens_in":16956,"tokens_out":2143,"would_cite":false,"duration_ms":17878,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Owners are whiter than residents in 81% of NY tracts","keywords":["racial equity","property ownership","race imputation","machine learning","census tracts","New York State","homeownership disparity","corporate ownership"],"falsifier":"Obtain true race/ethnicity for a random sample of individual property owners in majority-Black New York tracts, from self-identification in a survey, matched administrative records, or field verification, and compare observed owner shares with the model's predictions. If White ownership in those tracts is close to the 6.7% White population share rather than the predicted 39%, the central disparity estimate is largely a measurement artifact.","tokens_in":15989,"feed_emoji":"🏘️","tokens_out":8481,"duration_ms":66626,"temperature":0.7,"pith_summary":"This paper claims that property ownership in New York is racially skewed even after accounting for who lives in each neighborhood. Statewide, the average census tract has 5–10 percentage points more White owners than White residents, and in over 81% of tracts the predicted White owner share exceeds the White population share. The gap is steepest in minority-majority neighborhoods: in an average majority-Black tract, White individuals are estimated to own about 39% of properties although they are only 6.7% of residents. The paper reaches these numbers by applying a name- and geolocation-based machine-learning model to millions of parcel records, and it argues that corporate ownership widens the gap further by removing housing stock from minority owner-occupancy. If correct, the results map a racial wealth gap in property control at neighborhood scale and point to concrete housing-policy targets.","feed_headline":"Owners are whiter than residents in 81% of NY tracts","feed_subtitle":"Imputed owner race shows White ownership runs 5-10 points above White population, sharpest in minority-majority neighborhoods.","key_machinery":"The machine carrying the argument is a deep-learning race/ethnicity imputation system: a character-level LSTM neural network that reads owner names and, in the Full Model, also takes census-tract geolocation as input and passes outputs through an XGBoost filter, trained on labeled voter-registration data and previously validated at 89.2% accuracy. A Name-Only LSTM version is used for New York City, where owner addresses are missing. The model outputs per-owner probabilities over five race/ethnicity categories; the paper assigns the most likely category, aggregates ownership counts and assessed value by tract, and compares them with census population shares. A New York ground-truth sample of 35,134 PPP loan applicants who self-reported race gives the Full Model 87.25% accuracy and per-class false-positive rates, which the paper then applies to stress-test its extreme-disparity examples.","core_discovery":"The paper's central discovery is that New York property ownership is systematically whiter than the resident population at the census-tract level. Statewide, the mean tract is about 65% White in population but 72% White in ownership, and in more than 81% of tracts the predicted White owner share exceeds the White population share. The pattern is most dramatic in minority-majority tracts: in an average majority-Black tract, Black residents are 69.2% of the population but Black owners are only 44.6% of owners, while White residents are 6.7% of the population but White owners are 39.3%; majority-Hispanic tracts show Hispanic residents at 64.5% versus 38.3% Hispanic ownership and White residents at 10.2% versus 27.1% White ownership. White owners also hold about 87% of individually-owned assessed property value against roughly 75% of the state population. Corporate ownership widens the effective gap: in majority-Black, majority-Hispanic, and mixed tracts, White individual plus corporate ownership reaches 55.3%, 55.3%, and 56.7% respectively, meaning most housing stock in these communities is not held by the resident majority.","pith_inferences":["An extension the paper leaves implicit: the stress test corrects only four illustrative tracts; running the same error adjustments across all tracts would reveal whether the 81% overrepresentation figure and the 5–10 point average gap survive in aggregate.","Because corporate owners are not assigned a race, the 'White + corporate' combined shares rest on the untested assumption that corporate owners are effectively White-controlled; beneficial-ownership records would turn that assumption into a measurement.","The analysis does not separate White owner-occupants from absentee White landlords in minority tracts, so it cannot distinguish gentrification-driven displacement from long-run rent extraction; adding owner-address data would split those mechanisms.","Applied longitudinally to several years of assessment rolls, the same pipeline could track whether ownership gaps widen or narrow under new housing policies, a follow-up the paper itself identifies."],"forward_implications":["White overrepresentation is not a suburban or rural artifact; it persists inside Black- and Hispanic-majority tracts, so ownership and resident demographics diverge in the places where the resident majority is non-White.","Corporate ownership materially changes the disparity picture: in majority-Black, majority-Hispanic, and mixed tracts, combining White individual owners with corporate owners puts more than half of the housing stock under White-plus-institutional control.","The choice between Full and Name-Only models matters for New York City; since the Name-Only model is acknowledged to risk overestimating White ownership, the city's true disparities could be larger than the reported numbers.","Policy levers that follow from the analysis include tenant opportunity-to-purchase programs, restrictions on speculative LLC purchases, and support for minority homebuyers in tracts where ownership is most out of line with population."],"supporting_citations":[{"why":"Supplies the LSTM+Geo+XGBoost race/ethnicity imputation model and its bias-reduction validation, the methodological core of the study.","marker":"Chalavadi et al., 2025"},{"why":"Introduces Bayesian Improved Surname Geocoding, the baseline that the Full Model builds on and is compared against.","marker":"Elliott et al., 2009"},{"why":"Documents the misclassification bias in administrative-race imputation that motivates the Full Model's geolocation and filtering design.","marker":"Argyle and Barber, 2023"},{"why":"Provides PPP loan records with self-identified race, used as the ground-truth validation set and the source of error rates for the stress tests.","marker":"U.S. Small Business Administration, 2021a,b"},{"why":"Supplies the parcel-level property assessment rolls for properties outside New York City.","marker":"New York State, 2025"},{"why":"Supplies New York City property valuation and ownership records.","marker":"New York City Department of Finance, 2025"},{"why":"Frames corporate and institutional rental ownership as financialization of housing, which the paper uses to interpret its corporate ownership findings.","marker":"Fields and Uffer, 2016"},{"why":"Provides the financialization-of-housing theoretical framework cited when interpreting investor ownership in minority neighborhoods.","marker":"Aalbers, 2016"}],"fun_headline_variants":["NY owners whiter than residents in 81% of tracts","White owners hold 87% of NY property value, but 75% of population","In Black-majority NY tracts, White owners 39% vs 7% White residents","In minority-majority NY tracts, White and corporate ownership tops 55%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the race-imputation model, trained on voter rolls from Florida and North Carolina, transfers to New York property owners, and that error rates measured on PPP loan applicants apply directly to tract-level ownership predictions; if those error structures differ, the reported 5–10 point White overrepresentation could be partly a measurement artifact.","fun_headline_variants_meta":{"raw":{"variants":["NY owners whiter than residents in 81% of tracts","White owners hold 87% of NY property value, but 75% of population","In Black-majority NY tracts, White owners 39% vs 7% White residents","In minority-majority NY tracts, White and corporate ownership tops 55%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00209,"raw_usage":{"total_tokens":8178,"prompt_tokens":1051,"completion_tokens":7127,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":667,"completion_tokens_details":{"reasoning_tokens":7040}},"tokens_in":667,"tokens_out":7127,"duration_ms":41845,"temperature":1.0,"reasoning_tokens":7040,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:53:14.970603+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Obtain true race/ethnicity for a random sample of individual property owners in majority-Black New York tracts, from self-identification in a survey, matched administrative records, or field verification, and compare observed owner shares with the model's predictions. If White ownership in those tracts is close to the 6.7% White population share rather than the predicted 39%, the central disparity estimate is largely a measurement artifact.","supporting_citations":[{"cited_title":"STRATA: A Name-and-Geography Race Inference Model for Fair Lending and Housing Equity Applications","cited_arxiv_id":"2504.21259","evidence_quote":"Supplies the LSTM+Geo+XGBoost race/ethnicity imputation model and its bias-reduction validation, the methodological core of the study."},{"cited_title":"N., Morrison, P","cited_arxiv_id":null,"evidence_quote":"Introduces Bayesian Improved Surname Geocoding, the baseline that the Full Model builds on and is compared against."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents the misclassification bias in administrative-race imputation that motivates the Full Model's geolocation and filtering design."},{"cited_title":"and Uffer, S","cited_arxiv_id":null,"evidence_quote":"Frames corporate and institutional rental ownership as financialization of housing, which the paper uses to interpret its corporate ownership findings."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the financialization-of-housing theoretical framework cited when interpreting investor ownership in minority neighborhoods."}],"review_version":1}