Pith. sign in

REVIEW 3 major objections 4 minor 86 references

To what extent can long-differencing capture climate adaptation?

T0 review · 3 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Long-differencing and fixed-effects comparisons systematically understate climate adaptation, and the standard test for it is underpowered.

desk verdict Solid decomposition, overclaimed headline: the LD–FE gap is a shrunken multiple of the true adaptation gap, but the 30–80% 'understatement' is contradicted by the paper's own tables. read the letter →

arxiv 2607.22028 v1 pith:X53GLPQD submitted 2026-07-24 econ.EM

Dalia Ghanem , Felix Pretis , Daniel Schuurman This is my paper
classification econ.EM MSC 62P2091B76
keywords climateadaptationlongdifferencesfixedeffectscontaminationweighttestbyimplicationsignal-to-noiseratiopaneleconometrics
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper sets out to show that the common practice of measuring climate adaptation by comparing long-difference (LD) and fixed-effects (FE) estimates is fundamentally unreliable. It proves that neither estimator isolates its intended target: both estimands are weighted averages of the long-run and short-run responses to climate and weather. Because the LD–FE gap equals the true adaptation gap multiplied by a factor that is usually far below one, the measured gap is a shrunken version of the truth, and the standard null 'LD equals FE' is a test by implication that can have trivial power. An empirically calibrated simulation for U.S. corn yields finds the gap understates adaptation by roughly 30–80%, depending on the averaging window. A sympathetic reader would care because this calls into question a widely used method for estimating how much societies can adapt to climate change.

What carries the argument

The carrying object is the pair of contamination weights ω_S^LD and ω_L^FE, which measure the share of the LD-transformed regressor variation coming from weather shocks and the share of the FE-transformed variation coming from climate. Their sum controls the slope in Corollary 2.2. Because the within transformation removes most climate variation, ω_L^FE is tiny; because averaging reduces but does not eliminate weather variance, ω_S^LD is large, so the LD–FE gap is compressed. The paper also characterizes how the climate signal-to-noise ratio depends on the averaging window τ, showing that the bias-minimizing window is interior and depends on the unobserved climate process.

What would settle it

Generate a panel with known θ_S = -0.07 and θ_L = -0.02, climate defined as a 10-year normal from real temperature data, T = 25, τ = 5, and 1,000 replications. If the simulation mean of β̂_LD − β̂_FE does not equal (θ_L − θ_S)(1 − ω̂_S^LD − ω̂_L^FE), with ω̂_S^LD ≈ 0.825 and ω̂_L^FE ≈ -0.0265, then Corollary 2.2's decomposition is empirically wrong.

Watch

Extended reading notes

Core claim

The paper's central claim is that β_LD − β_FE = (θ_L − θ_S)(1 − ω_S^LD − ω_L^FE), where θ_L and θ_S are the long-run and short-run response parameters and the ω terms are contamination weights measuring how much of the wrong variation survives each transformation. Because both contamination weights are typically nonnegative in climate settings, the slope of the LD–FE comparison against true adaptation is less than one and can be zero when the weights sum to one. Thus the adaptation test based on the equality of the two estimators is a test by implication: equality is necessary but not sufficient for no adaptation. In simulations calibrated to U.S. temperature data, the slope is about 0.2 for

Load-bearing premise

The model is additively separable in latent climate and weather components, with weather shocks strictly exogenous conditional on the full history of observed weather; if the outcome is not additively separable, or if weather shocks are anticipated so strict exogeneity fails, the clean Corollary 2.2 decomposition no longer describes the estimands.

Editorial extensions

If this is right

  • Non-rejections of the LD–FE adaptation test cannot be interpreted as evidence of no adaptation; they are also consistent with adaptation masked by contamination.
  • Published LD–FE gaps are downward-biased measures of true adaptation; correcting them requires estimating the contamination weights under an explicit climate model.
  • The bias-minimizing averaging window τ depends on the unobserved climate process, so no single window is universally optimal.
  • Specifying a climate model permits construction of confidence intervals for θ_L − θ_S by test inversion, though weak identification arises when the contamination weights sum to one.
  • Even at the bias-minimizing window, the LD estimator can remain substantively biased when the climate signal-to-noise ratio is low.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the 30–80% downward bias carries over to real settings, many existing adaptation estimates in the literature are likely lower bounds; re-analysis using estimated contamination weights could shift the range of plausible adaptation.
  • The test-by-implication logic extends beyond LD–FE: any design that tests an implication of a null rather than the null itself needs a power analysis over the alternative space, not just size control.
  • The signal-to-noise decomposition suggests that panels with large interannual weather variability may require much longer time horizons or different estimators; researchers could use the estimated SNR to screen panel designs before running LD–FE comparisons.
  • Response heterogeneity, analyzed in Appendix C, makes the probability limits variance-weighted averages, so even a corrected LD–FE comparison no longer targets a single population adaptation parameter; heterogeneity in weather variability across units becomes a confound.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper studies the long-difference (LD) versus fixed-effects (FE) comparison that is widely used to infer climate adaptation. It posits an outcome model y_it = θ_S w*_it + θ_L c*_it + α_i + ε_it, with observed weather x_it = c*_it + w*_it, and derives the probability limits of FE and LD estimators. Propositions 2.1 and 2.2 show that each estimand is a weighted average of the short-run response θ_S and the long-run response θ_L. Corollary 2.2 then gives β_LD − β_FE = (θ_L − θ_S)(1 − ω_S^LD − ω_L^FE), so comparing LD and FE is a test by implication of the null θ_L = θ_S: equality of the estimands is necessary but not sufficient for no adaptation. The paper also analyzes how the LD contamination weight ω_S^LD depends on the averaging window τ through the climate signal-to-noise ratio, and reports an empirically calibrated simulation using Burke–Emerick temperature data. The abstract claims the LD–FE difference understates adaptation by about 30–80%.

Significance. The paper's core algebraic decomposition is a useful and clean formalization of an important identification concern. The weighted-average representations in Propositions 2.1 and 2.2 are proved transparently, and Corollary 2.2 gives a simple, interpretable expression for what LD–FE comparisons actually recover. The 'test by implication' logic is also valuable: a non-rejection of β_LD = β_FE does not establish the absence of adaptation. The analytical SNR results in Appendix B, and the explicit recognition that the bias-minimizing τ depends on the unobserved climate process, are genuine contributions. However, the paper's headline quantitative claim—that LD–FE understates adaptation by 30–80%—is not a robust implication of the analysis and is contradicted by several of the paper's own reported simulation cells. The heterogeneity extension in Appendix C, acknowledged in Remark 2.2, also undermines the size-control property of the adaptation test as stated. With the quantitative claims reframed and the homogeneity caveat made prominent, the paper has publishable value as an econometric cautionary analysis.

major comments (3)
  1. [Abstract and Section 2.6.2 / Table D3] The abstract claims that the LD–FE difference 'understates adaptation by about 30–80%', but this is contradicted by the paper's own Table D3. In the T=40, 10-year-normal, τ=20 row, the simulation means are ω̂_S^LD = −0.1249 and ω̂_L^FE = −0.0053, so Corollary 2.2 gives slope 1 − ω̂_S^LD − ω̂_L^FE = 1.1302. The LD–FE gap then overstates θ_L − θ_S by about 13%, not understates it. Conversely, Table D2 (Panel B, 30-year normal, τ=10) gives ω̂_S^LD = 0.9269 and ω̂_L^FE = −0.0055, implying a slope of about 0.078, i.e. an understatement of about 92%. The text on p. 14 does acknowledge that 'in other variants of our simulation design we find τ values that lead to over-estimation', but this is not reflected in the abstract or the introduction's summary. The 30–80% range is an artifact of selected calibration cells. The abstract and abstract-level conclusions must be reworded to describe the rang
  2. [Corollary 2.1, Table D3, and simulation reporting] Corollary 2.1's attenuation result requires the non-negative covariance assumptions E[Δw̄_i Δc̄_i] ≥ 0 and E[w̃_it c̃_it] ≥ 0. The reported simulation values violate this assumption: Table D3, Panel A, T=40, τ=20 reports ω̂_S^LD = −0.1249, which corresponds to a negative LD covariance contribution, and the FE contamination weight is also negative. The simulations are presented without flagging that they fall outside the maintained assumptions of Corollary 2.1. This matters because the negative weight is exactly what produces the sign reversal noted in the first comment. The tables should mark such cells, and the simulation discussion should explain whether these configurations are considered plausible or are boundary cases where the formal corollaries do not apply.
  3. [Remark 2.2 and Appendix C] The paper claims that the FE–LD comparison is a test by implication of no adaptation and therefore controls size. This claim rests on the homogeneous-response model y_it = θ_S w*_it + θ_L c*_it + α_i + ε_it. Under the heterogeneous-response model in Remark 2.2/Appendix C, y_it = θ_{S,i} w*_it + θ_{L,i} c*_it + α_i + ε_it, even when θ_{S,i} = θ_{L,i} = θ_i for every unit (no adaptation for any unit), β_FE and β_LD are variance-weighted averages of θ_i with different weights—within-transformed variation for FE versus long-differenced variation for LD. These weighted averages need not be equal. Thus H0: θ_L = θ_S does not generally imply β_LD = β_FE under heterogeneity, and the size-control property of the adaptation test fails. The paper acknowledges that the probability limits become variance-weighted averages, but it does not draw this implication for the test's validity. This is a load-
minor comments (4)
  1. [Table D3 note] The note to Table D3 says 'for T=25' but the table reports T=40 results. The note should read T=40.
  2. [Section 2.6.2, last paragraph] The text correctly acknowledges that some τ values overestimate θ_L − θ_S, but the wording around 'consistent with our formal results' and the abstract's 30–80% claim should be aligned with this acknowledgment.
  3. [Footnote 4] Typo: 'parantheses' should be 'parentheses'.
  4. [Section 3, Eq. (11)] The unit-root SNR expression is given as Tτ − 4τ²/3 + τ/3 − 1. It may be worth adding a sentence connecting this expression to the variance components in Proposition B.2 to help readers verify the simplification from the displayed covariance formulas.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the core algebra is derived from stated assumptions, and the headline magnitudes are conditional simulation outputs, not fitted predictions.

full rationale

The paper's central claims are derived from explicitly stated assumptions, not from the target result. Propositions 2.1 and 2.2 express β_FE and β_LD as weighted averages of θ_S and θ_L using the outcome model y_it = θ_S w*_it + θ_L c*_it + α_i + ε_it and x_it = c*_it + w*_it. Corollary 2.2 then follows by algebra (subtracting the two probability limits and using the fact that the weights sum to one). This is a parameter-free derivation that does not assume that LD-FE differences understate adaptation; the understatement conclusion instead follows when the contamination weights are positive, as in Corollary 2.1 under stated covariance assumptions. The empirically-calibrated simulation is an illustration, not a fitted prediction: θ_S and θ_L are inputs on a grid, c*_it is defined as a 10- or 30-year climate normal, and the reported 30-80% understatement is read off the simulated slopes, which are exactly (1 − ω_S^LD − ω_L^FE) by Corollary 2.2. This is a numerical consequence of the chosen DGP, not a prediction obtained by fitting the target. Self-citations (Carter et al. 2018, Ghanem and Smith 2021) are not load-bearing: they are cited as sources of a competing assumption or as related literature, and the proofs in the appendix stand on their own. The paper itself flags the key scope condition, the separable outcome model, and Appendix C shows the clean Corollary 2.2 slope is not robust to response heterogeneity; those are honest limitations and correctness concerns, not circularity. The concern that some reported cells in Table D3 produce slopes above one (so the 30-80% headline is not universal) is a robustness/interpretation issue, not a circularity issue.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central theorem rests on three domain assumptions: additive separable climate/weather decomposition, homogeneous responses, and strict exogeneity. The simulated numerical claims additionally depend on the chosen m, τ, and θ grid, which are hand-picked rather than estimated. No new particles, mediators, or invented objects are introduced.

free parameters (3)
  • Climate-normal window m = 10, 20, 30 years
    The simulation defines climate c* as the m-year trailing average of observed degree days (Section 2.6.1). The headline 30–80% understatement and even the sign of the LD bias depend on m.
  • LD averaging window τ = 5, 10, 20 in simulations
    τ is the estimator's tuning parameter. The paper shows ω_S^LD and the bias vary nonlinearly with τ, and the bias-minimizing τ depends on the unobserved climate process, so the specific numerical claims are conditional on the τ grid chosen.
  • Response parameters θ_S, θ_L grid = θ_S = −0.07; θ_L ∈ {−0.07, −0.06, ..., 0}
    The simulation calibrates θ_S to the Burke–Emerick FE estimate and varies θ_L to trace power. These are inputs, not estimated from the target result, but the reported percentages apply to this grid and not universally.
assumptions (4)
  • domain assumption Additive latent decomposition x_it = c*_it + w*_it, with c* interpreted as climate and w* as weather shock
    Introduced in Section 2.1, Eq. (2)–(3). All probability-limit results depend on this decomposition and on the outcome being separable in the two components.
  • domain assumption Outcome model y_it = θ_S w*_it + θ_L c*_it + α_i + ε_it with homogeneous responses θ_S, θ_L
    Eq. (3) is the load-bearing response model. Appendix C shows that the clean Corollary 2.2 decomposition does not survive even mild response heterogeneity without unit-specific variance weights.
  • domain assumption Strict exogeneity E[ε_it|X_i] = 0 for all i, t
    Section 2.3 states 'we impose strict exogeneity throughout the paper.' If shocks are anticipated and adaptation responses correlate with forecast errors, the estimands pick up additional channels beyond θ_L − θ_S.
  • domain assumption Nonnegative covariance between within- and LD-transformed climate and weather components
    Corollary 2.1's attenuation sign results require E[w̃c̃] ≥ 0 and E[Δw*Δc*] ≥ 0. The simulation's FE contamination weight is negative (ω_L^FE = −0.0265), so the 'attenuation' language in Section 2.6.2 is not literally covered by the corollary.

how reviews work

0 comments
Cite this review

Pith. "Pith review of To what extent can long-differencing capture climate adaptation?." pith.science (2026). https://pith.science/paper/X53GLPQD

@misc{pith2026260722028,
  author       = {Pith},
  title        = {Pith review of: To what extent can long-differencing capture climate adaptation?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/X53GLPQD}},
  note         = {Machine review of arXiv:2607.22028}
}
read the original abstract

Understanding the degree to which we are able to adapt to climate change is central to economic assessments of future climate damages. Economists increasingly use comparisons between long differences and fixed effects estimators to measure climate adaptation. We show that such comparisons can be misleading. Neither estimator is consistent for its intended parameter, as both the long-difference (LD) and fixed effects (FE) estimands are weighted averages of the long- and short-run responses to climate and weather. As a result, the difference between the two understates the true extent of adaptation, and the standard test based on this difference --while controlling size -- tends to be substantially underpowered in the settings researchers typically encounter. An empirically-calibrated simulation shows this difference understates adaptation by about 30--80%, depending on the averaging window.

Figures

Figures reproduced from arXiv: 2607.22028 by the authors.

Figure 1
Figure 1. Summary of choices of long-differencing averaging window [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Simulation Results: FE, LD and Adaptation Test [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗
Figure 3
Figure 3. Climate signal-to-noise ratio as a function of [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Sample estimates of SNRc across τ for different climate normals with T = 25 Notes: Panels (a) and (b) plot the sample analogues of E¯[∆¯c ∗2 i ] (Climate variation), E¯[∆ ¯w∗2 i ] (Weather variation) and E¯[∆¯c ∗ i ∆ ¯w∗ i ] (Covariance). Panel (c) and (d) present the …

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

86 extracted references · 1 canonical work pages

  1. [1]

    Journal of Environmental Economics and Management , volume=

    Climate Adaptive Response Estimation: Short and long run impacts of climate change on residential electricity and natural gas consumption , author=. Journal of Environmental Economics and Management , volume=. 2022 , publisher=

    2022

  2. [2]

    Journal of Economic Perspectives , volume=

    Quantifying economic damages from climate change , author=. Journal of Economic Perspectives , volume=. 2018 , publisher=

    2018

  3. [3]

    Journal of Political Economy , volume=

    Adapting to climate change: The remarkable decline in the US temperature-mortality relationship over the twentieth century , author=. Journal of Political Economy , volume=. 2016 , publisher=

    2016

  4. [4]

    Journal of Public Economics , volume=

    Climate and migration in the United States , author=. Journal of Public Economics , volume=. 2025 , publisher=

    2025

  5. [5]

    Empirical Economics , volume=

    The impact of drought on farm economic performance: evidence from Sweden , author=. Empirical Economics , volume=. 2025 , publisher=

    2025

  6. [6]

    Handbook of the Economics of Climate Change , volume=

    Adaptation to climate change , author=. Handbook of the Economics of Climate Change , volume=. 2024 , publisher=

    2024

  7. [7]

    Annual Review of Resource Economics , volume=

    Identifying the economic impacts of climate change on agriculture , author=. Annual Review of Resource Economics , volume=. 2018 , publisher=

    2018

  8. [8]

    Agriculture , author =

    Adaptation to Climate Change: New Evidence from U.S. Agriculture , author =. 2025 , howpublished =

    2025

Show all 86 references
  1. [9]

    Journal of Environmental Economics and Management , volume=

    A unifying approach to measuring climate change impacts and adaptation , author=. Journal of Environmental Economics and Management , volume=. 2023 , publisher=

    2023

  2. [10]

    local temperature , author=

    The macroeconomic impact of climate change: Global vs. local temperature , author=. 2024 , institution=

    2024

  3. [11]

    American Economic Journal: Economic Policy , volume=

    Adaptation to climate change: Evidence from US agriculture , author=. American Economic Journal: Economic Policy , volume=. 2016 , publisher=

    2016

  4. [12]

    2019 , institution=

    Climatic constraints on aggregate economic output , author=. 2019 , institution=

    2019

  5. [13]

    Nature Climate Change , volume=

    Adaptation of US maize to temperature variations , author=. Nature Climate Change , volume=. 2013 , publisher=

    2013

  6. [14]

    2024 , institution=

    Are We Adapting to Climate Change? , author=. 2024 , institution=

    2024

  7. [15]

    Proceedings of the national academy of sciences , volume=

    Crop-damaging temperatures increase suicide rates in India , author=. Proceedings of the national academy of sciences , volume=. 2017 , publisher=

    2017

  8. [16]

    The Quarterly Journal of Economics , volume=

    Valuing the global mortality consequences of climate change accounting for adaptation costs and benefits , author=. The Quarterly Journal of Economics , volume=. 2022 , publisher=

    2022

  9. [17]

    2024 , journal =

    The far-reaching distributional effects of global warming: Evidence from half a century of climate and inequality data , author=. 2024 , journal =

    2024

  10. [18]

    Journal of Development Economics , volume=

    Response and adaptation of agriculture to climate change: Evidence from China , author=. Journal of Development Economics , volume=. 2021 , publisher=

    2021

  11. [19]

    Journal of the Association of Environmental and Resource Economists , volume =

    Dalia Ghanem and Aaron Smith , title =. Journal of the Association of Environmental and Resource Economists , volume =. 2021 , doi =. https://www.journals.uchicago.edu/doi/pdf/10.1086/710968 , abstract =

    2021 doi

  12. [20]

    On model selection criteria for climate change impact studies , journal =

    Xiaomeng Cui and Bulat Gafarov and Dalia Ghanem and Todd Kuffner , keywords =. On model selection criteria for climate change impact studies , journal =. 2024 , note =. doi:https://doi.org/10.1016/j.jeconom.2023.105511 , url =

    2024

  13. [21]

    Econometrica , volume=

    Average and quantile effects in nonseparable panel models , author=. Econometrica , volume=. 2013 , publisher=

    2013

  14. [22]

    Journal of Economic Geography , pages=

    The autonomous adaptation of US homes to changing temperatures , author=. Journal of Economic Geography , pages=. 2025 , publisher=

    2025

  15. [23]

    Journal of Environmental Economics and Management , volume=

    Climate change and adaptation in agriculture: Evidence from US cropping patterns , author=. Journal of Environmental Economics and Management , volume=. 2020 , publisher=

    2020

  16. [24]

    Journal of Development Economics , volume=

    Climate change, cropland adjustments, and food security: Evidence from China , author=. Journal of Development Economics , volume=. 2024 , publisher=

    2024

  17. [25]

    Journal of Economic Surveys , volume=

    Does global warming worsen poverty and inequality? An updated review , author=. Journal of Economic Surveys , volume=. 2024 , publisher=

    2024

  18. [26]

    Nature Climate Change , pages=

    Impacts of global warming on subnational poverty and inequality , author=. Nature Climate Change , pages=. 2025 , publisher=

    2025

  19. [27]

    American economic review , volume=

    Bones, bombs, and break points: the geography of economic activity , author=. American economic review , volume=. 2002 , publisher=

    2002

  20. [28]

    American economic review , volume=

    Two-way fixed effects estimators with heterogeneous treatment effects , author=. American economic review , volume=. 2020 , publisher=

    2020

  21. [29]

    2005 , publisher=

    Optimal statistical decisions , author=. 2005 , publisher=

    2005

  22. [30]

    American Economic Journal: Macroeconomics , volume=

    Temperature shocks and economic growth: Evidence from the last half century , author=. American Economic Journal: Macroeconomics , volume=. 2012 , publisher=

    2012

  23. [31]

    Journal of Economic literature , volume=

    What do we learn from the weather? The new climate-economy literature , author=. Journal of Economic literature , volume=. 2014 , publisher=

    2014

  24. [32]

    2017 , institution=

    The marginal product of climate , author=. 2017 , institution=

    2017

  25. [33]

    Environmental Research Letters , volume=

    Historical warming has increased US crop insurance losses , author=. Environmental Research Letters , volume=. 2021 , publisher=

    2021

  26. [34]

    arXiv preprint arXiv:2505.17946 , year=

    The Effects of Climate and Weather on Economic Output: Evidence from Global Subnational Data , author=. arXiv preprint arXiv:2505.17946 , year=

    arXiv

  27. [35]

    Nature Climate Change , volume=

    A top-down approach to projecting market impacts of climate change , author=. Nature Climate Change , volume=. 2016 , publisher=

    2016

  28. [36]

    Journal of Environmental Management , volume=

    Climate change and mortality evolution in China , author=. Journal of Environmental Management , volume=. 2020 , publisher=

    2020

  29. [37]

    Journal of Econometric Methods , volume=

    Broken or fixed effects? , author=. Journal of Econometric Methods , volume=. 2019 , publisher=

    2019

  30. [38]

    arXiv preprint arXiv:2505.08950 , year=

    The Economic Impact of Low-and High-Frequency Temperature Changes , author=. arXiv preprint arXiv:2505.08950 , year=

    arXiv

  31. [39]

    Journal of the Association of Environmental and Resource Economists , volume=

    Temperature and human capital in the short and long run , author=. Journal of the Association of Environmental and Resource Economists , volume=. 2018 , publisher=

    2018

  32. [40]

    Review of Economics and Statistics , volume=

    Why you should never use the Hodrick-Prescott filter , author=. Review of Economics and Statistics , volume=. 2018 , publisher=

    2018

  33. [41]

    Review of Development Economics , volume=

    How climate change leads to emigration: Conditional and long-run effects , author=. Review of Development Economics , volume=. 2021 , publisher=

    2021

  34. [42]

    Review of Economics and Statistics , volume=

    Adaptation and the mortality effects of temperature across US climate regions , author=. Review of Economics and Statistics , volume=. 2021 , publisher=

    2021

  35. [43]

    American Economic Review , volume=

    The enduring impact of the American Dust Bowl: Short-and long-run adjustments to environmental catastrophe , author=. American Economic Review , volume=. 2012 , publisher=

    2012

  36. [44]

    Annual Review of Resource Economics , volume=

    Climate econometrics , author=. Annual Review of Resource Economics , volume=. 2016 , publisher=

    2016

  37. [45]

    Journal of Population Economics , volume=

    The impact of global warming on obesity , author=. Journal of Population Economics , volume=. 2024 , publisher=

    2024

  38. [46]

    Evidence from a global panel of regions , author=

    The impact of climate conditions on economic production. Evidence from a global panel of regions , author=. Journal of Environmental Economics and Management , volume=. 2020 , publisher=

    2020

  39. [47]

    Journal of Climate , volume=

    Separating internal variability from the externally forced climate response , author=. Journal of Climate , volume=

  40. [48]

    Theory , author=

    Stochastic climate models part I. Theory , author=. tellus , volume=. 1976 , publisher=

    1976

  41. [49]

    Environmental Research: Climate , volume=

    Origin, importance, and predictive limits of internal climate variability , author=. Environmental Research: Climate , volume=. 2023 , publisher=

    2023

  42. [50]

    Geophysical Research Letters , volume=

    ENSO change in climate projections: forced response or internal variability? , author=. Geophysical Research Letters , volume=. 2018 , publisher=

    2018

  43. [51]

    Nature , volume=

    Forcing, feedback and internal variability in global temperature trends , author=. Nature , volume=. 2015 , publisher=

    2015

  44. [52]

    Journal of Environmental Economics and management , volume=

    Adjustment costs from environmental change , author=. Journal of Environmental Economics and management , volume=. 2005 , publisher=

    2005

  45. [53]

    Review of Environmental Economics and Policy , year=

    Estimating the economic impacts of climate change using weather observations , author=. Review of Environmental Economics and Policy , year=

  46. [54]

    Agricultural Economics , volume=

    Distributional heterogeneity in climate change impacts and adaptation: Evidence from Indian agriculture , author=. Agricultural Economics , volume=. 2023 , publisher=

    2023

  47. [55]

    2018 , institution=

    Estimating the consequences of climate change from variation in weather , author=. 2018 , institution=

    2018

  48. [56]

    2025 , institution=

    A guide to climate damages , author=. 2025 , institution=

    2025

  49. [57]

    American Economic Journal: Economic Policy , volume=

    Climate change and labor reallocation: Evidence from six decades of the Indian Census , author=. American Economic Journal: Economic Policy , volume=. 2023 , publisher=

    2023

  50. [58]

    Review of Environmental Economics and Policy , year=

    Measuring climate adaptation: Methods and evidence , author=. Review of Environmental Economics and Policy , year=

  51. [59]

    Energy Economics , volume=

    Is temperature adversely related to economic development? Evidence on the short-run and the long-run links from sub-national data , author=. Energy Economics , volume=. 2024 , publisher=

    2024

  52. [60]

    American Journal of Agricultural Economics , volume=

    Climate econometrics: Can the panel approach account for long-run adaptation? , author=. American Journal of Agricultural Economics , volume=. 2021 , publisher=

    2021

  53. [61]

    Journal of Environmental Economics and Management , volume=

    Sufficient statistics for climate change counterfactuals , author=. Journal of Environmental Economics and Management , volume=. 2024 , publisher=

    2024

  54. [62]

    Environmental research letters , volume=

    Negative impacts of climate change on cereal yields: statistical evidence from France , author=. Environmental research letters , volume=. 2017 , publisher=

    2017

  55. [63]

    Journal of development Economics , volume=

    The long-run impact of bombing Vietnam , author=. Journal of development Economics , volume=. 2011 , publisher=

    2011

  56. [64]

    Oxford Open Economics , volume=

    Climate change and economic activity: evidence from US states , author=. Oxford Open Economics , volume=. 2023 , publisher=

    2023

  57. [65]

    Nature Climate Change , volume=

    Adaptation potential of European agriculture in response to climate change , author=. Nature Climate Change , volume=. 2014 , publisher=

    2014

  58. [66]

    The Quarterly Journal of Economics , volume=

    Does directed innovation mitigate climate damage? Evidence from US agriculture , author=. The Quarterly Journal of Economics , volume=. 2023 , publisher=

    2023

  59. [67]

    Journal of Political Economy , volume=

    Climate change, the food problem, and the challenge of adaptation through sectoral reallocation , author=. Journal of Political Economy , volume=. 2025 , publisher=

    2025

  60. [68]

    2024 , publisher=

    How much will global warming cool global growth? , author=. 2024 , publisher=

    2024

  61. [69]

    2024 , institution=

    Migration, Climate Similarity, and the Consequences of Climate Mismatch , author=. 2024 , institution=

    2024

  62. [70]

    Proceedings of the National Academy of sciences , volume=

    Adapting North American wheat production to climatic challenges, 1839--2009 , author=. Proceedings of the National Academy of sciences , volume=. 2011 , publisher=

    2009

  63. [71]

    Proceedings of the National Academy of Sciences , volume=

    Empirical evidence of mental health risks posed by climate change , author=. Proceedings of the National Academy of Sciences , volume=. 2018 , publisher=

    2018

  64. [72]

    Nature communications , volume=

    Using insurance data to quantify the multidimensional impacts of warming temperatures on yield risk , author=. Nature communications , volume=. 2020 , publisher=

    2020

  65. [73]

    NBER Working Paper , number=

    Temperature, adaptation, and local industry concentration , author=. NBER Working Paper , number=

  66. [74]

    World Bank Working Paper , year=

    Heat, Informality, and Misallocation , author=. World Bank Working Paper , year=

  67. [75]

    Proceedings of the National Academy of sciences , volume=

    Nonlinear temperature effects indicate severe damages to US crop yields under climate change , author=. Proceedings of the National Academy of sciences , volume=. 2009 , publisher=

    2009

  68. [76]

    Food Policy , volume=

    Extreme weather events cause significant crop yield losses at the farm level in German agriculture , author=. Food Policy , volume=. 2022 , publisher=

    2022

  69. [77]

    Available at SSRN 3212073 , year=

    Improving climate damage estimates by accounting for adaptation , author=. Available at SSRN 3212073 , year=

  70. [78]

    Proceedings of the National Academy of sciences , volume=

    The fingerprint of climate trends on European crop yields , author=. Proceedings of the National Academy of sciences , volume=. 2015 , publisher=

    2015

  71. [79]

    Climatic Change , volume=

    Climate change and violent conflict in Europe over the last millennium , author=. Climatic Change , volume=. 2010 , publisher=

    2010

  72. [80]

    Journal of the Association of Environmental and Resource Economists , volume=

    Irrigation and climate change: Long-run adaptation and its externalities , author=. Journal of the Association of Environmental and Resource Economists , volume=. 2026 , publisher=

    2026

  73. [81]

    WMO Guidelines on the Calculation of Climate Normals , institution =

  74. [82]

    European Review of Agricultural Economics , volume=

    Warming temperatures, yield risk and crop insurance participation , author=. European Review of Agricultural Economics , volume=. 2021 , publisher=

    2021

  75. [83]

    American Journal of Agricultural Economics , volume=

    Understanding the effect of cover crop use on prevented planting losses , author=. American Journal of Agricultural Economics , volume=. 2024 , publisher=

    2024

  76. [84]

    Environmental Research Letters , volume=

    Rising temperature threatens China’s cropland , author=. Environmental Research Letters , volume=. 2022 , publisher=

    2022

  77. [85]

    Scientific reports , volume=

    Maladaptation of US corn and soybeans to a changing climate , author=. Scientific reports , volume=. 2021 , publisher=

    2021

  78. [86]

    Journal of econometrics , volume=

    Errors in variables in panel data , author=. Journal of econometrics , volume=. 1986 , publisher=

    1986

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.