Research and Editorial Methods

The research question determines the unit of analysis, period and evidence required. A study states what it seeks to explain and identifies its contribution: a finding, a comparison, a policy assessment or a clarification of measurement. Breadth of coverage is not a substitute for an answer.

The U.S. economy provides the main reference point. International research examines the institutions and exposures relevant to a comparison, including cases in which the American experience cannot be transferred directly.

Primary statistical releases, official accounts, legal instruments and policy documents establish the source record. Research literature is used for methods, empirical findings and competing explanations. A directory identifies an issuer or dataset; a substantive claim requires the relevant document, table or provision.

Sources are cited near the claims they support. Source text, published data, analytical inference and hypothetical illustration are distinguished. A source may support one proposition without supporting every conclusion in a paragraph.

Quantitative work identifies the series, units, geography, reference period, price basis, adjustment and release vintage where these affect interpretation. Rates, shares, levels and contributions are treated as different measures.

Currency conversion, annualization, index construction and ratios use an explicit formula or stated procedure. Illustrative calculations disclose their assumptions and are not mixed with observed data. Rounding is applied after the underlying calculation; small differences caused by rounding are identified when material.

A comparison requires a common basis in coverage, definition and time. Where reconciliation is incomplete, the remaining difference is part of the finding. Institutional differences are examined before treating a cross-country relationship as a policy result.

Causal language requires a credible account of the counterfactual and the influences that could otherwise explain the result. Descriptive evidence can be informative without identifying causation. Scenarios describe conditional outcomes and do not acquire the status of forecasts through numerical precision.

Policy analysis distinguishes proposals, adopted measures, appropriations, implementing decisions, operational programs and evaluations. The relevant legal or administrative source establishes the scope and timing of an obligation. Announcements and explanatory documents supply context within their stated authority.

An allocation is not equivalent to an expenditure, and expenditure is not equivalent to an outcome. Analysis follows the administrative and economic stages through which a measure changes behavior. Recommendations state the objectives and tradeoffs on which they depend.

Internal review examines source correspondence, arithmetic, policy scope, reasoning, terminology and the consistency of the conclusion with the evidence. The publication identifies its information cutoff. This is an internal editorial assessment; source agencies remain responsible for their own publications.

A material factual or analytical error warrants correction and an explanation sufficient to identify the affected conclusion. Routine copy changes should not obscure a substantive revision. Readers may submit a documented observation through the corrections process.

Readers should cite the publication title, AES, the public address and the information cutoff or relevant edition. Source agencies and original authors retain credit for their work. Permission to read a source does not establish permission to republish its tables, images or other protected material.

The source guide describes the roles of official and multilateral material. Terms of use addresses reproduction of website content and third-party rights.