Evidence (562 claims)
Search and filter individual claims pulled from the papers. Looking for a specific finding ("what's the effect on wages?"), you're in the right place. Want to compare whole outcome categories against each other instead? Use the Evidence Explorer.
The board below groups claims two ways: by broad theme (nine paper-level topics) and by outcome category (the 34 claim-level outcomes that the Explorer and Syntheses also use).
Browse by theme
Nine broad, paper-level topics. Click one to filter the claims below.
Adoption
21267 claims
Filter claims →
Productivity
17978 claims
Filter claims →
Governance
17038 claims
Filter claims →
Human-AI Collaboration
16914 claims
Filter claims →
Org Design
11104 claims
Filter claims →
Innovation
11087 claims
Filter claims →
Labor Markets
6711 claims
Filter claims →
Skills & Training
5616 claims
Filter claims →
Inequality
4343 claims
Filter claims →
Claims by outcome category
Counts by direction of finding. These are the same 34 outcome categories the Explorer compares and the Syntheses are written for. A linked row has a published synthesis.
| Outcome | Positive | Negative | Mixed | Null | Total |
|---|---|---|---|---|---|
| Other | 1880 | 496 | 296 | 1854 | 4721 |
| Organizational Efficiency | 2906 | 665 | 438 | 180 | 4210 |
| Governance & Regulation | 2162 | 929 | 480 | 247 | 3866 |
| Technology Adoption Rate | 1533 | 545 | 278 | 210 | 2593 |
| Decision Quality | 1391 | 534 | 321 | 173 | 2429 |
| Output Quality | 1298 | 472 | 231 | 145 | 2153 |
| AI Safety & Ethics | 682 | 821 | 230 | 90 | 1837 |
| Research Productivity | 855 | 253 | 121 | 425 | 1675 |
| Firm Productivity | 1105 | 171 | 175 | 73 | 1531 |
| Task Allocation | 735 | 229 | 361 | 99 | 1433 |
| Market Structure | 457 | 461 | 251 | 47 | 1222 |
| Innovation Output | 673 | 94 | 108 | 36 | 913 |
| Task Completion Time | 499 | 118 | 43 | 38 | 702 |
| Firm Revenue | 458 | 130 | 61 | 26 | 677 |
| Skill Acquisition | 381 | 122 | 113 | 34 | 650 |
| Consumer Welfare | 316 | 176 | 115 | 39 | 648 |
| Employment Level | 223 | 143 | 177 | 53 | 600 |
| Error Rate | 246 | 282 | 44 | 19 | 594 |
| Fiscal & Macroeconomic | 283 | 142 | 78 | 52 | 562 |
| Inequality Measures | 103 | 329 | 106 | 13 | 552 |
| Worker Satisfaction | 225 | 185 | 63 | 30 | 503 |
| Automation Exposure | 158 | 155 | 72 | 37 | 426 |
| Regulatory Compliance | 186 | 126 | 35 | 14 | 362 |
| Team Performance | 193 | 56 | 51 | 24 | 326 |
| Developer Productivity | 224 | 58 | 27 | 13 | 323 |
| Wages & Compensation | 148 | 108 | 50 | 17 | 323 |
| Training Effectiveness | 218 | 44 | 21 | 27 | 313 |
| Job Displacement | 23 | 159 | 53 | 5 | 240 |
| Hiring & Recruitment | 109 | 61 | 32 | 11 | 215 |
| Skill Obsolescence | 16 | 107 | 26 | 6 | 155 |
| Creative Output | 71 | 44 | 28 | 6 | 150 |
| Social Protection | 58 | 31 | 12 | 3 | 104 |
| Labor Share of Income | 29 | 43 | 25 | 2 | 99 |
| Worker Turnover | 45 | 29 | 6 | 4 | 84 |
| Industry | — | — | — | 1 | 1 |
The same Sub-Saharan African panel study found that tax-system efficiency supports higher revenue, while compliance burdens can reduce collection performance.
Panel study of ten Sub-Saharan African countries.
MAGA and tax-incidence narratives attenuate the average expectation revision without materially reducing dispersion, while changing the causal explanations agents provide.
Comparison of T1, T8, and T9, which vary the narrative attached to an otherwise similar immediate 10 percent tariff message; open-ended responses are used to examine explanations.
Message complexity affects average expectation revisions: minimalist wording produces a larger average revision than standard wording, while technical language produces a smaller revision; both unusually sparse and unusually complex messages widen dispersion relative to standard wording.
Comparison of treatment arms T1, T6, and T7, varying semantic complexity while keeping the tariff and implementation timing constant.
A progressively escalating sequence of tariff messages produces a larger and more persistent expectation response than a single high-rate message, while repeated semantic reversals weaken the response.
Comparison of T4, T10, and T11, which hold the terminal message fixed while varying the preceding sequence between a single message, reversals, and progression.
Differences in digital-services taxation and transfer-pricing rules affect where AI economic value is captured and can produce tax-planning distortions or double taxation that influence AI-firm location decisions.
Qualitative policy analysis applying the study's comparative taxation analysis to digital and AI services; no quantitative estimate of location or tax effects is provided.
Institutional theory, agency theory, and fiscal federalism provide complementary explanations for local-authority debt accumulation: persistence and routinization of debt-producing practices, accountability failures, and intergovernmental incentives, respectively.
Theoretical synthesis of three frameworks used to organize the literature.
The proposed framework treats institutional drivers as antecedents to debt, governance mechanisms as moderators of debt outcomes, and councils as context-specific units whose scale shapes debt dynamics.
Theoretical synthesis combining institutional theory, agency theory, and fiscal federalism; this is a proposed conceptual framework rather than an empirically estimated model.
In the authors' illustrative macroeconomic exercise, the currently automatable share implies negligible aggregate effects of around 0.1 percentage points per year, Wave 1 implies about 0.9 percentage points, Wave 2 around 4 percentage points, and Wave 3 more than 20 percentage points in annual productivity and price effects.
Order-of-magnitude calculation assuming displaced employment is fully automated over roughly ten years at an even rate, with displaced labor redeployed; the authors explicitly state that these are not forecasts.
Digital transformation reduces total carbon emissions only when production-digitization investment efficiency is sufficiently high.
Theoretical carbon-emissions analysis under different production-digitization investment-efficiency conditions within the carbon cap-and-trade model.
AI in education may increase aggregate human capital and long-run productivity, but unequal access may widen income and opportunity gaps.
The paper's economic implications synthesis, drawing on the cross-national review and documented implementation evidence; no quantified macroeconomic estimates are reported.
The emissions impact of additional data-center electricity demand depends on the marginal generation mix, transmission constraints, and the location and timing of demand.
Conceptual and literature-based assessment distinguishing average from marginal emission factors and citing electricity-system studies.
China’s higher education spending remained comparatively restrained while its universities rose in global rankings, increasing from approximately 1.0% of GDP in the early 2000s to about 1.4% by the early 2020s.
Descriptive comparison of Chinese higher education expenditure as a share of GDP over time and against OECD and United States benchmarks.
The paper proposes a theoretical transmission mechanism in which AI adoption and task substitution lead to cognitive substitution, marginal-cost compression, price and income compression, fiscal pressure, and institutional responses.
The supplied conceptual-architecture description summarizes the sequence developed in the paper. It is a theoretical framework rather than an empirical estimate, and no data or sample are reported.
AI-enabled education and innovation capacity can support economic growth, but sustainability benefits are conditional on complementary energy, governance, and climate policies.
Interpretation of the empirical results: positive long-run association between AI readiness/STEM capacity and economic modernization, combined with negative association between AI readiness and climate transition; policy conclusion drawn from these combined findings and dynamic analyses (VECM, impulse responses).
The tax system is becoming not only a fiscal but also a technological institution, directly impacting the state's economic security.
Argumentative claim in the paper supported by analysis of FTS statistics (2019–2024) and institutional discussion linking digitalization/AI adoption to economic-security implications.
The positive macroeconomic benefits of AI are dwarfed by persistent structural issues in the economy.
Comparative assessment in the study: quantitative modelling results contrasted with analysis of structural constraints; no numerical comparison provided in the excerpt.
If labor displacement becomes the dominant trend under AI, we could see a sustained joint increase in both unemployment and output.
Counterfactual/theoretical scenario presented by the authors (analytical reasoning about possible simultaneous rise in unemployment and output); no empirical sample or quantified evidence in the excerpt.
The effects of including own-account data as an asset differ significantly across industries.
Industry-level results from the modified ILPA showing heterogeneous impacts across industries when own-account data are capitalized and capital services are allocated to data-using industries.
The impact of AI investments on economic development (HDI) varies across countries — i.e., there is parameter heterogeneity across the panel.
Results of the Pesaran–Yamagata homogeneity test and heterogeneous-panel CS-ARDL estimation applied to the 8-country 2012–2023 panel.
The study examines how AI affects total factor productivity (TFP) and GDP growth.
Model-based analysis reported in the paper (task-based economic model assessing macroeconomic outcomes). No empirical sample size reported.
Macro-level productivity effects remain uneven due to slow diffusion, disparities in digital readiness, and the need for complementary organizational and human-capital investments.
Review-level synthesis drawing on macro and cross-firm studies and theoretical arguments presented in the paper (2010–2025 literature).
The different movements suggest that markets may anticipate open and closed AI advances to have different economic implications.
Interpretation by the author based on observed opposite yield movements for open versus closed model releases; inference about market expectations. This is an interpretive claim rather than a direct measurement. Underlying empirical basis: the opposite-direction yield shifts reported above.
Patterns are similar for treasuries, corporate bonds, and TIPS.
Author reports that the same opposite-direction pattern across open vs closed models is observed for multiple bond types (treasuries, corporate bonds, and TIPS). No quantitative results or sample counts included in the excerpt.
Long-term bond yields shift in opposite directions following the introduction of open versus closed AI models.
Author's empirical result from the extended analysis comparing market responses to open versus closed model releases. Method implied: comparison/event-study of bond yields around release dates for open and closed models. Specific sample sizes, estimation details, and statistical significance not provided in the excerpt.
Trade openness (TO) positively influences sustainable economic growth (SEG) in China but has a weaker effect on SEG in the United States.
Country-specific ARDL results using quarterly data (2015Q1–2025Q4, 44 observations) showing a positive and meaningful TO coefficient for China and a smaller (weaker) coefficient for the United States.
Long-run asymmetric response to renewable energy shocks is statistically confirmed (Wald χ² = 5.42, p = 0.020).
Long-run Wald test for asymmetry from CS-PMG-NARDL on the 18-country panel (2000–2023); reported χ² and p-value.
Short-run asymmetric response to renewable energy shocks is statistically confirmed (Wald χ² = 4.102, p = 0.043).
Short-run Wald test for asymmetry from CS-PMG-NARDL on the 18-country panel (2000–2023); reported χ² and p-value.
These positive results are not supported in all contexts (i.e., the positive effects are not universally found across all specifications/contexts).
Abstract statement noting heterogeneity/robustness: preferred results hold but are 'not supported in all contexts.' Implies some specifications or subsets do not show the effects.
A 2025 forecasting study of experts reveals an apparent disconnect between expectations of significant AI capability improvements and modest near-term economic projections.
2025 forecasting study / expert elicitation involving 69 leading economists and 52 AI experts, plus additional expert panels; comparison of experts' expectations about AI capability progress versus their near-term economic projections.
Empirical evidence remains heterogeneous, and estimates of AI’s macroeconomic contribution vary across institutional and structural contexts.
Synthesis of heterogeneous empirical studies from the 2015–2025 literature identified in the structured review; comparative thematic classification highlighting variation by institutional/structural context.
AI adoption does not generate uniform or automatic growth effects.
Structured literature review / mechanism-oriented synthesis covering studies from 2015–2025; transparent search, screening and thematic classification (no formal meta-analysis).
A forecast-error variance decomposition attributes most aggregate volatility to the longevity shock, while the AI shock dominates the variance of the return to AI capital.
Model-based forecast-error variance decomposition implemented on the simulated stochastic model to apportion variance of aggregate variables and the return to AI capital across shocks.
The macroeconomic significance of AI-induced productivity depends not only on technological efficiency, but also on the distributive transmission of productivity gains through labour income, disposable income, prices, investment, public expenditure, transfers and external demand.
Theoretical argument and synthesis of literature in the conceptual review (no new empirical estimation reported).
With endogenous capital accumulation, data-driven automation generates explosive growth but stagnant long-run wages.
Extended model incorporating endogenous capital accumulation: analytical solution/characterization showing unbounded (explosive) growth in aggregate variables while real wages remain stagnant in the long run (model derivation).
Aggregate AI metrics (the composite AI Vibrancy Score) obscure heterogeneous pillar-level effects on tourism’s economic contribution.
Comparison of null result for the aggregate AI Vibrancy Score with significant positive effects for specific pillars (R&D, Policy and Governance, lagged Talent) in the same fixed-effects analyses on 33 countries (2017–2023).
There is a long-run equilibrium (cointegrating) relationship among AI adoption, skill-disaggregated unemployment, and sustainable development in South Africa.
Empirical ARDL results reported in the paper indicating a long-run equilibrium relationship based on annual 2003–2024 time-series data.
The paper examines the macroeconomic impact of AI (drawing on the cited institutional projections) to understand sectoral and aggregate economic implications for Georgia.
Method: macroeconomic synthesis of external projections (Goldman Sachs, McKinsey, Penn Wharton, IMF) and application to Georgia; no reported experimental sample size.
Resource (digital talent) agglomeration should remain at a moderate level and achieve coordinated development, because excessive concentration can reduce the growth benefits (implied by the inverted-U finding).
Policy implication drawn from the paper’s finding of an inverted-U relationship between talent agglomeration, industrial digitalization, and regional economic growth (normative recommendation based on empirical nonlinear result).
The relation among digital talent agglomeration, industrial digitalization, and regional economic growth follows an inverted-U shape (consistent with the Williamson hypothesis).
Systematic empirical examination of China's provincial regions using regional-level empirical analysis (paper reports an econometric test of nonlinear/quadratic relationships between digital talent agglomeration, industrial digitalization, and regional economic growth). Sample size (number of provinces/observations) not stated in the excerpt.
The negative quadratic term confirms a concave (inverted-U) relationship between AI and economic growth (diminishing marginal returns of AI).
Panel data for 19 G20 countries (2005–2023) estimated with a quadratic specification in GMM; reported negative and statistically significant coefficient on the AI-squared term.
Even when two economies share the same long-run automation level, adoption speed alone determines transition welfare.
Comparative-welfare analysis in the dynamic theoretical model holding long-run automation level fixed while varying adoption speed (analytical comparative statics).
Modeling fiscal policy as a government problem (instead of an abstract planner) implies a tax changes the firm's automation first-order condition, raises revenue only on the remaining automation base, and requires specifying rebates and administrative losses.
Explicit governmental optimization and budget-accounting setup in the model: taxes enter firms' automation first-order conditions; revenue is computed on post-tax automation activity and rebates/administration are modeled.
The magnitude of AI’s effect on potential GDP varied across industries and depended on the level of digital maturity, human resources, and institutional conditions.
Decompositional analysis across aggregated industry data and scenario-based modeling drawing on sectoral sources and reviews.
If employment losses are relatively small and productivity gains are realised, AI adoption could boost Exchequer revenues. But if job displacement is sizeable, tax receipts fall while welfare spending rises, resulting in potentially large pressures on the public finances.
Conditional fiscal scenarios simulated in the report combining employment, wage and benefit changes with the public finance implications (tax receipts and welfare spending); reported as scenario-based outcomes.
Automation leads economic growth to accelerate, but the acceleration is remarkably slow because of the prominence of 'weak links' (an elasticity of substitution among tasks substantially less than one); even when most tasks are automated by rapidly-improving capital, output is constrained by the tasks performed by slowly-improving labor.
Theoretical mechanism from the task-based model (σ < 1 weak-links structure) combined with calibrated simulations that incorporate historical accounting results.
The growth effects of AI are conditional on institutional quality and organizational adaptability.
Theoretical/analytical claim in the paper's framework and supported by the stylized-facts analysis indicating heterogeneity in productivity and growth outcomes by institutional and digital capacity indicators.
The magnitude and timing of macroeconomic impact from quantum computing are highly uncertain.
Monte Carlo / scenario ensemble results showing wide (fat-tailed) outcome distributions driven by uncertainty in technical milestones, adoption rates, and complementarity strengths; use of expert elicitation to parameterize tail risks.
Artificial intelligence (AI) has a positive and statistically significant effect on growth at lower conditional quantiles (τ = 0.10–0.25) but is insignificant at higher quantiles.
MMQR estimation results reported in the paper showing significant positive AI coefficients at τ = 0.10–0.25 and insignificant coefficients at higher quantiles.
Kondratieff, Schumpeter, and Mandel each highlight different drivers of capitalist long waves: Kondratieff emphasizes regular technological-driven renewal, Schumpeter emphasizes entrepreneurship and innovation-led creative destruction, and Mandel emphasizes class relations and production structures.
Comparative theoretical analysis and literature synthesis across the three schools; conceptual summary of canonical positions (no original dataset; qualitative interpretation).
Bringing motorcycle transport into the tax net in Ebonyi State did not produce strong revenue outcomes and was accompanied by low remittances and indications of avoidance or revenue leakage.
Evidence from a study of motorcycle transport taxation in Ebonyi State; the paper does not report the study's sample size in the supplied text.