Economics Research: Systematic Literature Review

Apply Claude Projects to economics research. Conduct a systematic literature review on 'AI's impact on labor markets' with 25 papers, extracting research questions, methods, findings, and gaps.

Economics Research Use Case

Conduct a systematic literature review on "AI's Impact on Labor Markets" using Claude Projects to analyze 25 economics papers from top journals and working paper series.

Research Goal: Synthesize findings on how artificial intelligence affects employment, wages, skill premiums, and labor market polarization.

Target Output: 4,000-word literature review with thematic analysis, identified research gaps, and methodological synthesis.

Economics-Specific Advantage: Claude Projects excels at extracting econometric methodologies, identifying instrumental variables, comparing identification strategies across studies, and detecting inconsistencies in empirical findings. This capability transforms literature review from a manual annotation task into a systematic, reproducible research process.

Paper Selection Strategy

Identify Core Paper Sources

Primary sources for high-quality economics research:

Tier 1: NBER Working Papers, AER, QJE, Econometrica, JPE, ReStud
Tier 2: Labour Economics, Journal of Labor Economics, ILR Review
Tier 3: ArXiv preprints, CEPR Discussion Papers, IZA Working Papers

Target 25 papers published 2018-present focusing on AI and labor markets.

Download Papers in PDF Format

Search strategy:

Google Scholar: "artificial intelligence" + "labor market" + "employment"
NBER Database: Filter by JEL codes J23 (Labor Demand), O33 (Technological Change)
ArXiv: cs.CY (Computers and Society) + econ.GN (General Economics)

Download all papers to a local folder for upload to Claude Projects.

Upload Papers to Claude Project

Create project named "AI Labor Markets Literature Review" and upload all 25 PDFs.

Verify uploads:

  • Check file names are descriptive (author-year format)
  • Confirm all papers are readable (OCR quality)
  • Organize by publication type if needed

Extraction Workflow

Extract Research Questions and Hypotheses

Prompt Claude to identify core research questions:

Extract: (1) Main research question (2) Testable hypotheses
(3) Contribution to literature (4) Relationship to prior work

Expected output: Structured list of RQs with page references.

Identify Econometric Methodologies

Extraction prompt for methods:

Extract: (1) Econometric methodology (DiD, IV, RDD, etc.)
(2) Data sources and sample characteristics (3) Identification strategy
(4) Robustness checks performed

Focus on comparing identification strategies across papers.

Capture Key Findings with Effect Sizes

Economics-specific extraction:

Extract: (1) Main findings with point estimates (2) Confidence intervals
(3) Statistical significance levels (4) Economic significance interpretation
(5) Heterogeneous effects by subgroup

Create table format for cross-study comparison.

Document Limitations and Threats

Critical analysis extraction:

Extract: (1) Acknowledged limitations (2) Threats to identification
(3) External validity concerns (4) Data limitations
(5) Suggestions for future research

This reveals research gaps systematically.

Synthesis Strategy

Identify Thematic Clusters

Ask Claude to group papers by theme:

  • Automation and job displacement (7-10 papers)
  • Skill-biased technological change (5-7 papers)
  • Wage effects and inequality (4-6 papers)
  • Labor market polarization (3-5 papers)
  • Policy responses and interventions (2-4 papers)

Generate thematic summary for each cluster.

Compare Methodological Approaches

Synthesize across identification strategies:

Compare: DiD studies vs IV studies vs structural models
Question: Do findings converge despite different methods?
Analyze: Which identification strategies are most credible?

Create methodology comparison table.

Detect Contradictions and Inconsistencies

Prompt for critical analysis:

Identify: (1) Contradictory findings across studies (2) Measurement differences
(3) Sample selection issues (4) Time period variations
Explain: Why might results differ?

This reveals where field lacks consensus.

Map Research Gaps

Systematic gap identification:

Analyze: (1) Underexplored contexts (developing economies, specific industries)
(2) Methodological gaps (lack of RCTs, limited long-run studies)
(3) Theoretical gaps (mechanisms not tested) (4) Data limitations

Output: Prioritized list of 5-7 research gaps for future work.

Output Example Structure

Claude Projects can generate this outline with detailed summaries:

Literature Review Structure (4,000 words)

1. Introduction (500 words)

  • Research motivation and policy relevance
  • Scope definition and inclusion criteria
  • Review organization overview

2. Theoretical Framework (600 words)

  • Task-based models (Acemoglu & Restrepo)
  • Skill-biased technological change theory
  • Labor market polarization mechanisms

3. Empirical Evidence: Job Displacement (800 words)

  • Automation and routine task substitution
  • Cross-country comparison of displacement rates
  • Industry-specific findings (manufacturing, services)

4. Empirical Evidence: Wage Effects (700 words)

  • Wage premiums for AI-complementary skills
  • Inequality implications across skill groups
  • Geographic variation in wage impacts

5. Methodological Synthesis (600 words)

  • Identification strategy comparison
  • Data sources and measurement approaches
  • Robustness of findings across methods

6. Research Gaps and Future Directions (500 words)

  • Underexplored contexts and populations
  • Methodological improvements needed
  • Policy-relevant research priorities

7. Conclusion (300 words)

  • Consensus findings and open questions
  • Implications for policy and practice

Economics-Specific Tips

Econometric Terminology Precision: When extracting methodologies, verify Claude correctly identifies difference-in-differences versus event study designs, instrumental variables versus control function approaches, and fixed effects versus random effects models. Economics has precise terminology that matters for interpretation. Always cross-reference extracted methods with paper's actual specification tables.

Citation Management: Claude can extract APA7 citations from PDFs, but verify author names, publication years, and journal titles against official sources. For working papers, note version dates and update status. Create a separate bibliography document within the project to track citation metadata systematically.

Verification Checklist

After completing synthesis, verify:

  • All 25 papers analyzed and cited
  • Research questions extracted for each paper
  • Methodologies correctly identified (DiD, IV, RDD, etc.)
  • Key findings include point estimates and significance levels
  • Thematic clusters logically organized
  • Contradictions explicitly addressed
  • Research gaps prioritized by importance
  • Literature review outline has clear narrative flow
  • Citations formatted consistently (APA7)
  • Cross-references between papers identified

Expected Time Savings: Manual literature review of 25 papers typically requires 40-60 hours (reading, annotating, synthesizing). With Claude Projects, reduce this to 8-12 hours (uploading, prompt engineering, verification). The AI handles exhaustive extraction while you focus on critical analysis and narrative construction.

Next Steps