Economics Dissertation Topics
BSc, MSc & PhD Research Ideas
The definitive resource for economics dissertation planning — covering 100+ carefully developed research topics across macroeconomics, microeconomics, development economics, behavioural economics, environmental and ecological economics, financial economics, labour economics, health economics, international trade, and political economy. Each topic comes with research angles, identification strategies, data sources, and methodology guidance for BSc, MSc, and PhD students at every stage of dissertation development.
📊 Need expert support with your economics dissertation? Our specialists cover every sub-field and level.
Get Expert Help →What Is an Economics Dissertation — and Why Does Your Topic Choice Shape Everything?
An economics dissertation is a substantial, independently conducted research project that applies economic theory, empirical methods, or both to an original question in a defined area of economic enquiry. It is the capstone of every economics degree — the assessment in which students move from consuming economics to producing it. Whether you are writing 10,000 words at BSc level, 20,000 at MSc level, or a three-paper thesis at PhD level, the quality of your research question, your identification strategy, and your relationship between theory and evidence will determine the quality of the final product more decisively than any other factor. This guide maps the full landscape of modern economics dissertation research, providing 100+ specific, analytically grounded topic ideas across ten major sub-fields — with the methodology, data source, and thesis-formulation guidance that transforms a broad research area into a credible dissertation project.
Here is the scenario every economics dissertation student recognises: you have identified a broad area — inequality, climate policy, financial crises, labour markets — and you sit down to write a research proposal, only to discover that you cannot articulate a specific question, a specific dataset, or a specific methodology. The problem is not a lack of interest or economics knowledge. It is a misunderstanding of what an economics research question requires: not a topic, but a puzzle — a specific empirical regularity, theoretical tension, or policy question whose answer is not already obvious and whose investigation is feasible within your time and data constraints.
The distinction between a research area and a research question is the most important conceptual move in economics dissertation planning. “Income inequality” is a research area. “Does the introduction of a minimum wage increase reduce household poverty rates in low-wage UK regions, using a difference-in-differences strategy exploiting variation in regional minimum wage uptake?” is a research question — and the gap between those two statements is precisely the gap this guide is designed to help you close. For expert support at every stage of dissertation development — from topic selection and proposal writing to data analysis and final editing — the specialist team at Smart Academic Writing offers dissertation writing services, data analysis and statistics support, and thesis coaching for economics students at every level.
Theory vs. Empirical vs. Mixed-Methods Economics Dissertations
Theoretical economics dissertations develop, extend, or critique formal models — using mathematical economics to derive new predictions or generalisations about economic behaviour. They are more common at PhD level and in programmes with a strong theory emphasis. Empirical economics dissertations use data — survey data, administrative records, natural experiments, RCT findings — to test hypotheses, estimate causal effects, or document economic phenomena. They are the dominant form at BSc and MSc level and require a clear identification strategy. Mixed-methods dissertations combine formal modelling with empirical estimation or qualitative fieldwork — more common in development economics and political economy. Knowing which type your programme expects, and which your question demands, is the essential first step before any topic is selected.
BSc, MSc & PhD Economics Dissertations: What Each Level Demands
The three levels of economics dissertation are genuinely different intellectual projects — not simply longer and shorter versions of the same task. The expectations around methodological rigour, originality of contribution, depth of literature engagement, and sophistication of identification strategy differ substantially across levels. Understanding those differences is essential to scoping your topic, choosing your methods, and framing your contribution appropriately.
BSc Economics Dissertation
Focused empirical investigation using established economic methods and publicly available data sources
- Addresses a specific, focused research question in a defined sub-field
- Uses one primary empirical method (OLS, DiD, event study) applied correctly
- Engages substantively with the core secondary literature but need not be exhaustive
- Uses publicly available data from ONS, World Bank, FRED, or similar sources
- Demonstrates understanding of the key assumptions behind chosen methods
- Does not need to produce a genuinely novel contribution to frontier knowledge
- Evaluated on question quality, method appropriateness, evidence, and interpretation
- Supervisor guidance typically 4–6 meetings across the academic year
MSc Economics Dissertation
Rigorous empirical research with a credible identification strategy and systematic engagement with the frontier literature
- Requires a clearly specified identification strategy for causal questions
- Systematic literature review mapping the sub-field’s major contributions and gaps
- Must engage with robustness checks, sensitivity analysis, and alternative specifications
- Often uses administrative data, proprietary datasets, or merged public datasets
- Expected to make a modest empirical contribution — new context, new data, or extended method
- Should engage critically with existing literature, not just summarise it
- Evaluated against the question: does this advance understanding of the topic?
- Often forms the basis of a future working paper or publication
PhD Economics Thesis
Three to four original, publishable papers making a frontier contribution to economic knowledge
- Each chapter must make an original, publishable contribution to a specific literature
- Requires genuinely novel data, identification strategy, or theoretical contribution
- Comprehensive engagement with the frontier literature in each sub-field addressed
- Chapters often presented at seminars, workshops, and conferences before submission
- Evaluated against the standard: is each paper publishable in a top-field journal?
- Requires development of a distinctive research programme with internal coherence
- Supervised intensively with weekly or fortnightly meetings throughout
- Culminates in an oral examination (viva voce) with expert external examiners
How to Choose Your Economics Dissertation Topic: A Strategic Research Framework
Choosing an economics dissertation topic is a process of progressive refinement — from a broad sub-field through an interesting empirical puzzle to a specific, answerable research question with a credible identification strategy and a viable data source. Most students begin with a policy domain or economic phenomenon that captured their attention in a taught module, then work backward from “what would I need to show?” to identify whether the question is tractable given available data and methods. The most common mistake is choosing a question that is interesting but unanswerable — either because no suitable data exists, because the identification strategy is too demanding for the level of the project, or because the question requires theoretical machinery not yet available in the literature.
The Four-Circle Test for a Strong Economics Dissertation Topic
The best economics dissertation topics sit at the intersection of four circles: (1) genuine intellectual interest — you will spend months with this question, and intrinsic motivation matters more than any other factor in sustaining dissertation quality; (2) data availability — the research question must be answerable with data you can actually access within your timeline and budget; (3) a credible identification strategy — for causal questions, there must be a plausible source of exogenous variation; and (4) a genuine gap in the literature — the question must not have already been definitively answered. If your topic is missing any one of these four elements, the dissertation will struggle at one or more stages. Choose the question, then find the data — but verify the data exists before committing to the question.
The Economics Dissertation Topic Development Timeline
Identify a Sub-Field and Policy Domain
Begin with the area of economics that most engaged you in your taught modules — labour markets, monetary policy, development, environmental economics, or any other sub-field. Read two or three recent review articles or handbook chapters in that area to understand what the key debates and frontier questions currently are. The American Economic Review’s Papers and Proceedings and the Journal of Economic Perspectives are ideal starting points — both publish accessible, frontier-orienting pieces aimed at the broader economics community.
Map the Frontier Literature Using EconLit
Search EconLit (the AEA’s bibliographic database) and the NBER Working Paper Series for recent publications in your chosen area. Identify the two or three most-cited recent papers in the domain, read them carefully, and pay specific attention to their “further research” sections — these are explicit invitations to subsequent researchers. The gaps that established scholars identify are the most legitimate and tractable gaps for dissertation research at any level.
Identify Data Sources and Assess Feasibility
Before formulating your research question, verify that the data required to answer it actually exists and is accessible to you. Check the World Bank Open Data portal, FRED, ONS, UK Data Service, and any sector-specific administrative datasets relevant to your question. Assess whether the required variation for your identification strategy is present in the data. A technically strong question that cannot be answered with available data is not a viable dissertation topic at any level.
Formulate a Specific Research Question and Identification Strategy
Convert your area of interest into a specific research question that specifies the outcome variable, the treatment or independent variable of interest, the population, the time period, and the identification strategy. “What is the causal effect of X on Y in population Z, identified using variation W?” is the template. This level of specificity is what separates a dissertation topic from a research area — and it is what your supervisor will expect to see in your proposal.
Write the Research Proposal
A 1,500–2,500 word proposal articulating your research question, its contribution to the existing literature, your data sources, your identification strategy, and your anticipated results. This is a commitment to a specific intellectual project — and it should be written with enough precision that your supervisor can give you genuinely useful feedback about whether the identification strategy is credible, the data is sufficient, and the contribution is real. For support developing your economics dissertation proposal, our thesis coaching service and dissertation writing service are available at BSc, MSc, and PhD level.
Macroeconomics Dissertation Topics: Growth, Cycles, Policy & Monetary Economics
Macroeconomics — the study of aggregate economic behaviour, including national output, inflation, unemployment, monetary policy, fiscal policy, and long-run growth — remains one of the most intellectually fertile and policy-relevant areas of economic research. Macroeconomics dissertation topics range from the highly empirical (estimating fiscal multipliers using instrumental variables) to the more theoretically ambitious (extending the New Keynesian model to incorporate financial frictions). The key challenge for macroeconomics dissertations at BSc and MSc level is the identification problem: macro time series are short, the variables of interest are highly collinear, and exogenous variation is rare. The strongest macro dissertations address this problem explicitly — either by using cross-country variation, exploiting natural experiments or policy discontinuities, or clearly acknowledging the limitations of the identification strategy used.
Macroeconomics & Monetary Policy Dissertation Topics
Growth theory, business cycles, fiscal and monetary policy, inflation, and central banking
Fiscal Multipliers in the Zero Lower Bound: Does Austerity Contract Output More During Recessions?
Estimates of the government spending multiplier vary sharply across business cycle states, monetary regimes, and country characteristics. This dissertation uses cross-country panel data and instrumental variables — military spending as an instrument for government expenditure in the tradition of Ramey (2011) — to estimate state-dependent fiscal multipliers and test whether fiscal consolidation is more contractionary when the nominal interest rate is at its lower bound.
Data: IMF World Economic Outlook; OECD Government at a Glance; Jordà-Schularick-Taylor Macrohistory Database. Method: Panel LP-IV (local projections with instrumental variables). BSc: OLS panel estimation; MSc/PhD: full IV local projection with bootstrap standard errors and state-dependent specifications.Quantitative Easing and Asset Price Inflation: Did Central Bank Asset Purchases Widen Wealth Inequality?
Following the 2008 financial crisis and COVID-19 pandemic, central banks in the UK, US, eurozone, and Japan purchased trillions of dollars in government bonds and mortgage-backed securities. This dissertation examines the distributional consequences of QE through its effects on equity prices, bond yields, and house prices — assessing whether the wealth effects disproportionately benefited asset-owning households.
Data: Bank of England Quarterly Bulletin; Federal Reserve Flow of Funds; ONS Wealth and Assets Survey; HFCS (Household Finance and Consumption Survey). Method: Event study around QE announcement dates; SVAR with sign restrictions for causal identification. Relevant for MSc and PhD level given data complexity.Inflation Persistence and Expectations De-anchoring: Is the 2020s Inflation Episode Different?
The post-2021 inflation surge in advanced economies prompted debate about whether inflation expectations had become de-anchored from central bank targets for the first time since the 1970s. This dissertation estimates measures of inflation persistence using time-varying parameter VAR models and tests whether the post-pandemic inflation episode exhibits structural break characteristics consistent with expectations de-anchoring.
Data: Federal Reserve Bank of Cleveland inflation expectations; BIS consumer expectation surveys; Consensus Economics forecasts. Method: TVP-VAR (time-varying parameter VAR) or Markov-switching model. Best suited for MSc dissertations with strong econometrics background.The China Shock Revisited: Import Competition, Regional Employment, and Political Polarisation
Acemoglu, Autor, Dorn, Hanson, and Price’s (2016) “China Shock” paper found large negative local labour market effects from Chinese import competition in the US, with subsequent work linking these effects to political polarisation. This dissertation replicates the core analysis for a European country (UK, Germany, or France) using the shift-share (Bartik) instrumental variables approach and tests whether the effects on employment and wages hold in the European institutional context.
Data: UN Comtrade; Eurostat Labour Force Survey; NUTS-2 regional data. Method: Shift-share IV (Bartik instrument) following Autor et al. Excellent MSc topic with a clear replication-and-extension design.Debt Sustainability and Sovereign Default Risk in Emerging Markets: What Macro Fundamentals Predict Spreads?
Sovereign default episodes — Argentina (2001, 2019), Greece (2012), Sri Lanka (2022) — raise fundamental questions about the macro conditions under which default risk becomes self-fulfilling and what fiscal and monetary fundamentals best predict sovereign spread movements. This dissertation builds a panel model of sovereign spreads for emerging market economies and tests the predictive power of fiscal balances, external debt, reserves, and growth forecasts.
Data: IMF World Economic Outlook; JP Morgan EMBI spreads; World Bank Global Debt Database. Method: Panel fixed effects; dynamic panel GMM; quantile regression for tail risk estimation. Accessible BSc topic with panel OLS; extendable to GMM for MSc.The Macroeconomics Identification Problem: A Critical Note
Macroeconomic research faces a fundamental challenge that microeconomic and development economics researchers have largely addressed: most macro time series are short (50–80 years of annual data), highly persistent, and characterised by few independent observations. This makes causal identification using macro data alone extremely difficult. The best BSc and MSc macroeconomics dissertations address this by using cross-country panel data (many countries over many years, giving genuine cross-sectional variation), exploiting natural experiments at the macro level (a specific policy change in one country during a specific period), or clearly acknowledging the descriptive — rather than causal — nature of their findings. Dissertations that run OLS on two macro time series and claim to have established causality are the most common and most serious methodological error in undergraduate and MSc macroeconomics research.
Microeconomics Dissertation Topics: Markets, Incentives, Information & Industrial Organisation
Microeconomics — the study of individual decision-making, market structure, incentives, information asymmetries, and game-theoretic interactions — offers some of the most tractable and methodologically exciting dissertation topics available to economics students. Because microeconomics questions are often answered using firm-level, household-level, or individual-level data, the identification challenges are more addressable than in macroeconomics, and the credible identification toolkit (RDD, DiD, IV, RCT) maps naturally onto micro questions. Industrial organisation, market design, and information economics are particularly active areas with rich dissertation potential at all three academic levels.
Platform Competition and Market Tipping in Digital Markets
How winner-take-all dynamics emerge in platform markets with strong network effects; empirical evidence on market concentration in digital advertising, ride-hailing, and food delivery; the conditions under which multi-homing prevents tipping; policy implications of market design interventions including data portability and interoperability requirements in the context of the EU Digital Markets Act and UK CMA digital markets frameworks.
Adverse Selection in Insurance Markets: Evidence from Health and Auto Insurance
Testing for adverse selection using the positive correlation test (Chiappori and Salanié, 2000) in health insurance, auto insurance, or annuity markets; how information asymmetries between insurer and insured generate welfare losses and market unravelling; the role of mandatory coverage, risk pooling, and genetic information regulation in managing adverse selection in modern insurance markets.
Tax Incidence and Distributional Effects of Commodity Taxes: VAT Pass-Through and Consumer Burden
Estimating the degree to which VAT changes are passed through to consumer prices, using scanner data or price index data around tax reform events; the distributional consequences of VAT given differential budget shares across the income distribution; evidence from UK VAT changes, EU harmonisation, and temporary VAT reductions during cost-of-living crises — with implications for the design of progressive tax systems.
Collusion and Cartel Detection: Using Structural Models to Identify Coordinated Pricing in Oligopolistic Markets
Structural econometric models of oligopolistic competition (Bertrand, Cournot, conduct parameter estimation) applied to detect departures from competitive pricing consistent with collusion; applications to pharmaceutical markets, airline pricing, petrol retail, or supermarket chains; the role of algorithmic pricing in facilitating tacit collusion without explicit communication — a frontier question in digital competition economics that has generated intense regulatory attention from the UK CMA, EU DG COMP, and US DOJ Antitrust Division.
Performance Pay, Incentive Design, and Worker Productivity: Evidence from Field Experiments
The design of optimal incentive contracts under moral hazard; empirical evidence from field experiments on how performance pay schemes (piece rates, tournaments, bonus structures) affect worker effort, selection, and productivity; the role of non-monetary incentives (recognition, intrinsic motivation) in domains where output is multi-dimensional and measurability is imperfect — with applications to education, healthcare, and public sector employment where standard performance pay approaches produce documented distortions.
Development Economics Dissertation Topics: Poverty, Inequality, Institutions & Randomised Trials
Development economics — which studies the economic conditions of low- and middle-income countries, the causes and consequences of poverty, and the design and evaluation of development interventions — has been transformed over the past two decades by the credibility revolution in empirical economics. The widespread adoption of randomised controlled trials (RCTs), difference-in-differences, and regression discontinuity designs has made it possible to establish credible causal estimates of the effects of microfinance, conditional cash transfers, educational interventions, health programmes, and institutional reforms. Development economics now offers some of the richest dissertation possibilities in all of economics — with public datasets from J-PAL, IPA, the World Bank, and DHS providing accessible empirical foundations at every level of study.
Development Economics Dissertation Topics
Poverty, institutions, education, health, gender, and the evaluation of development interventions
Conditional Cash Transfers and Human Capital Accumulation: A Meta-Analysis of RCT Evidence
Conditional cash transfer programmes — including Mexico’s PROGRESA/Oportunidades, Brazil’s Bolsa Família, and Kenya’s GiveDirectly — condition income support on school attendance, health check-up compliance, or nutritional monitoring. This dissertation conducts a systematic meta-analysis of the RCT literature on CCT effects on school enrolment, test scores, and long-run earnings, addressing heterogeneity across conditionality structures, transfer amounts, and country contexts.
Data: J-PAL’s Dataverse of published RCTs; World Bank DIME datasets; published study-level effect sizes. Method: Meta-analytic regression (WLS with inverse-variance weighting); publication bias tests (funnel plots, Egger’s test). Excellent BSc topic using published summary statistics; full micro-data analysis for MSc.Microfinance, Credit Access, and Household Welfare: Does Financial Inclusion Reduce Poverty?
The early enthusiasm for microfinance as a development panacea has been tempered by a series of rigorous RCT evaluations — in India (Banerjee et al., 2015), Ethiopia, Mexico, Mongolia, and Morocco — that found more modest effects than the field’s advocates claimed. This dissertation replicates and extends one of these landmark studies using publicly available replication data, or conducts a meta-analysis of the microfinance RCT literature to identify the conditions under which credit access generates measurable welfare improvements.
Data: Replication datasets from Banerjee et al. (2015) available via J-PAL; Demographic and Health Surveys; FinScope household surveys. Method: ITT/TOT estimation from RCT; IV using lottery randomisation as instrument for take-up. Strong MSc topic with clear methodological foundation.Colonial Institutions and Long-Run Economic Development: The Persistence of Extractive vs. Inclusive Institutions
Acemoglu, Johnson, and Robinson’s (2001) settler mortality instrument opened a vast literature on how colonial institutional choices — extractive vs. inclusive — have persisted as development determinants centuries later. This dissertation tests the AJR framework using updated institutional data, alternative instruments, or a specific regional or country-level case study, examining channels through which historical institutions affect contemporary productivity, education, and governance.
Data: Acemoglu-Johnson-Robinson replication data; V-Dem institutional quality database; Maddison Project for historical GDP; Colonial Database. Method: IV regression with settler mortality as instrument; RDD around colonial boundaries (following Michalopoulos and Papaioannou). A PhD-level literature applied at MSc with a bounded scope.Gender Gaps in Labour Markets and Entrepreneurship in Sub-Saharan Africa: Barriers and Interventions
Persistent gender gaps in employment, wages, entrepreneurship, and asset ownership in Sub-Saharan Africa reflect both formal institutional barriers (property rights, legal standing) and informal norms. This dissertation uses DHS or World Bank Enterprise Survey data to document the gender gaps and — exploiting policy variation across countries and time — tests whether specific legal reforms (collateral rights, business registration reform) or social norm interventions close measurable portions of those gaps.
Data: Demographic and Health Surveys (DHS); World Bank Enterprise Survey; World Bank Women, Business and the Law database. Method: DiD exploiting variation in legal reform timing across countries; decomposition analysis (Oaxaca-Blinder). Well-suited to MSc with cross-country panel.Aid Effectiveness and Fungibility: Does Foreign Aid Increase Investment or Substitute for Government Revenue?
The debate over foreign aid effectiveness — whether official development assistance raises investment, growth, and welfare, or whether it is partly fungible, diverted to consumption, or distorting of political incentives — has generated a vast empirical literature with contested findings. This dissertation uses panel data for aid-recipient countries and exploits donor-side variation in aid allocation (UN General Assembly voting patterns as an instrument, following Dreher et al.) to estimate the causal effect of aid on investment, fiscal revenue, and GDP growth.
Data: AidData (aiddata.org); World Bank World Development Indicators; IMF Article IV Consultation data; OECD DAC aid statistics. Method: Panel IV using political proximity to major donors as aid instrument. This is a classic and technically rigorous MSc or PhD topic.Behavioural Economics Dissertation Topics: Biases, Nudges, Decision-Making & Welfare
Behavioural economics — which incorporates insights from psychology into economic models of decision-making, challenging the standard assumptions of full rationality, time-consistency, and well-defined stable preferences — has grown from a provocative sub-discipline into one of the most policy-relevant and empirically active areas in the entire economics profession. The work of Kahneman, Thaler, Ariely, and their collaborators has established that systematic departures from standard rationality — present bias, loss aversion, overconfidence, anchoring, and framing effects — are robust, predictable, and consequential for economic outcomes including savings, health behaviour, retirement planning, tax compliance, and energy use. Behavioural economics dissertation topics are among the most accessible and practically impactful available to students at all three levels.
Default Options and Organ Donation: Does Opt-Out Save More Lives?
Comparing organ donation rates across countries with opt-in vs. opt-out default registration systems; exploiting Wales’s 2015 shift to presumed consent as a natural experiment to estimate the causal effect of default-framing on donation rates, with implications for libertarian paternalism and the design of public health nudges.
Procrastination, Commitment Devices, and Retirement Savings
Testing whether auto-enrolment in pension schemes (exploiting the UK’s staged auto-enrolment rollout 2012–2018) increases retirement savings beyond what present-biased workers would voluntarily choose; the role of commitment devices and contribution escalation in sustaining savings; heterogeneous treatment effects by income, age, and financial literacy.
Loss Aversion in Consumer Markets: Pricing, Reference Points, and the Endowment Effect
Testing for loss aversion in real consumer settings — housing markets, financial decisions, or labour supply — where the standard model’s predictions diverge from loss-averse models; using survey experiments or scanner data to estimate the curvature of the value function around reference points and the magnitude of the loss aversion coefficient.
Social Comparisons, Norms, and Energy Conservation: Field Experiment Evidence
Replicating or extending the Opower/Oracle field experiments that use social norm messaging (“your neighbours use 15% less electricity than you”) to reduce household energy consumption; estimating the effect size, persistence, and heterogeneity of norm-based nudges across household types and energy price levels — with direct policy relevance for carbon reduction without carbon pricing.
Overconfidence in Financial Markets: Evidence from Retail Investor Trading Data
Barber and Odean’s (2001) classic finding that overconfident investors trade too much and earn lower risk-adjusted returns has generated a large empirical literature. This dissertation uses retail broker trading records (where accessible) or survey-experimental data to test whether overconfidence — measured by miscalibration in return forecasts or excessive trading frequency — predicts portfolio underperformance, with attention to gender differences in overconfidence and their financial consequences.
Present Bias and Preventive Healthcare Uptake: Evidence from Low-Income Settings
The demand for preventive health products (insecticide-treated bed nets, water purification tablets, vaccines) in low-income settings is often well below socially optimal levels. This dissertation tests whether present-biased time preferences predict low uptake — using either an existing RCT dataset or survey-based elicitation of discount rates — and whether commitment mechanisms or small-time subsidies shift behaviour more than large unconditional subsidies at the point of decision.
Humans are not irrational — they are predictably irrational. Behavioural economics does not replace standard economics; it enriches it by identifying the systematic patterns in human decision-making that depart from the rational agent model in ways that are robust, replicable, and policy-relevant.
— A synthesis of the foundational argument in Thaler and Sunstein’s Nudge (2008) and Kahneman’s Thinking, Fast and Slow (2011)Environmental Economics Dissertation Topics: Carbon Pricing, Climate Policy & Green Transition
Environmental and ecological economics has become one of the most intellectually exciting and policy-urgent sub-fields in the discipline, driven by the climate and biodiversity crises and by the growing recognition that standard economic models systematically under-account for environmental externalities. The economics of climate change — carbon pricing mechanisms, the social cost of carbon, green technology adoption, and the distributional effects of environmental policy — offers dissertation topics of the highest academic and policy importance. Natural experiments abound: the EU Emissions Trading Scheme, national carbon taxes in Canada, Sweden, and the UK, and regional air quality regulations provide credible sources of quasi-experimental variation for causal identification.
Environmental & Climate Economics Dissertation Topics
Carbon markets, green transition, biodiversity economics, and the distributional effects of environmental policy
The EU Emissions Trading Scheme and Firm-Level Emissions: Has Carbon Pricing Reduced Industrial CO₂?
The EU ETS — the world’s largest carbon market — has been operational since 2005, generating a rich panel dataset of firm-level emissions, permit prices, and production outcomes. This dissertation uses DiD or synthetic control methods to estimate the causal effect of ETS coverage on firm-level emissions intensity, testing whether the scheme has reduced emissions beyond business-as-usual trends and identifying heterogeneous effects by sector, firm size, and permit allocation regime across Phases I–IV.
Data: EU Transaction Log (EUTL); Eurostat EU ETS data viewer; ORBIS firm-level financials. Method: DiD exploiting variation in ETS coverage across firms and sectors; synthetic control for aggregate sector-level analysis. Excellent MSc topic with freely available EUTL data.Carbon Taxes and Carbon Leakage: Do Stringent Domestic Policies Simply Export Emissions?
A major critique of unilateral carbon pricing is that it relocates emissions-intensive production to less-regulated jurisdictions rather than reducing global emissions — so-called carbon leakage. This dissertation estimates carbon leakage rates using bilateral trade data and energy intensity statistics, testing whether the introduction of the EU Carbon Border Adjustment Mechanism (CBAM) is consistent with leakage theory and examining whether leakage rates vary systematically with trade exposure and industry carbon intensity.
Data: UN Comtrade; EXIOBASE multi-regional input-output tables; EU ETS EUTL; IEA Energy Statistics. Method: Panel IV using carbon price variation as treatment; gravity model for trade flows. Strong MSc or PhD research design.The Distributional Effects of Green Energy Transition: Who Bears the Burden of Carbon Taxes?
Carbon taxes are often criticised as regressive because energy expenditure constitutes a larger share of lower-income household budgets. This dissertation uses expenditure survey data to estimate the distributional incidence of a carbon price across the income distribution, decomposing effects into direct (energy bills) and indirect (consumer goods price increases) channels, and assessing whether revenue-recycling mechanisms (dividend payments, income tax cuts, targeted subsidies) can offset the regressive incidence.
Data: ONS Living Costs and Food Survey; EU Household Budget Surveys; COICOP expenditure categories. Method: Input-output price model for indirect burden calculation; incidence analysis by decile. Excellent BSc topic with clear policy relevance and accessible data.Air Pollution and Cognitive Performance: Evidence from School Test Score Data
A growing body of research — including Lavy, Ebenstein, and Roth (2014) and Ebenstein et al. (2016) — documents significant negative effects of short-run air pollution exposure on cognitive performance, measured by test scores, standardised exams, and productivity data. This dissertation exploits day-to-day or week-to-week variation in air quality (PM2.5 or NO₂ concentrations) around school exam dates, using wind direction as an instrument for local pollution, to estimate the causal effect of pollution on academic performance.
Data: DEFRA UK Air Quality Network; Ofsted school data; KS2/KS4 results. Method: IV using wind direction instrument for local pollution variation. A strong, creative MSc topic with a compelling policy narrative about urban air quality and educational inequality.Financial Economics Dissertation Topics: Asset Pricing, Banking, Markets & Risk
Financial economics — the application of economic theory to financial markets, asset pricing, corporate finance, banking, and risk management — offers dissertation topics at the cutting edge of both academic research and real-world policy. The 2008 Global Financial Crisis, the COVID-19 market shock of 2020, and the more recent volatility around cryptocurrency markets, central bank digital currencies, and climate-related financial risk have generated a rich set of unresolved empirical questions with high policy stakes. Financial economics dissertations at MSc and PhD level typically require access to commercial financial databases (Bloomberg, WRDS, Datastream, Compustat), though several high-quality topics are feasible using freely available data from FRED, Yahoo Finance, or central bank statistical publications.
| Research Topic | Key Question & Identification | Level |
|---|---|---|
| ESG Investing and Stock Returns: Does Sustainability Generate Alpha or Impose a Costs? | Whether environmental, social, and governance (ESG) ratings predict risk-adjusted returns; using ESG score changes as treatment and four-factor model alphas as outcome; testing for mispricing vs. preference-driven pricing of sustainability risk | BSc/MSc |
| Cryptocurrency Market Efficiency: Are Crypto Prices Predictable? | Testing weak-form efficiency in Bitcoin, Ethereum, and altcoin markets using variance ratio tests, auto-correlation analysis, and GARCH modelling; assessing whether crypto markets have become more informationally efficient over time as institutional participation has grown | BSc/MSc |
| Bank Capital Requirements, Credit Provision, and the Real Economy | Whether increases in bank capital requirements (Basel III implementation) reduce credit supply and depress investment and output; exploiting variation in implementation timing across jurisdictions using DiD; Khwaja-Mian loan-level approach where firm-bank matched data available | MSc/PhD |
| Fintech Lending and Financial Inclusion: Do Platform Lenders Reduce Credit Exclusion? | Whether fintech lenders (Funding Circle, LendingClub, Kabbage) extend credit to borrowers excluded by traditional banks; using credit registry data to test whether fintech approvals predict lower default rates (adverse selection) or better served creditworthy borrowers | MSc |
| Insider Trading Detection: Market Microstructure Signals Before Corporate Announcements | Documenting abnormal order flow, bid-ask spread widening, and options market activity in the days before major corporate announcements (M&A, earnings, regulatory approvals); whether these patterns are consistent with informed trading and their implications for market surveillance | BSc/MSc |
| Central Bank Digital Currencies and Financial Stability: A Theoretical and Comparative Analysis | How CBDC introduction could alter the bank funding model through retail deposit substitution; comparison of the People’s Bank of China digital yuan pilot data with ECB consultation responses; implications for the monetary transmission mechanism and bank runs in a CBDC world | MSc/PhD |
| Climate-Related Financial Risk: Are Banks Adequately Pricing Transition and Physical Risk? | Using loan-level data or bank portfolio disclosures to test whether financial institutions have appropriately priced the credit risk associated with high-carbon assets and climate-exposed real estate; NGFS scenario analysis applied to UK banking sector exposures | MSc/PhD |
Labour Economics Dissertation Topics: Wages, Employment, Inequality & the Future of Work
Labour economics — the study of labour markets, wage determination, employment, unemployment, human capital, and labour market inequality — is one of the sub-fields most responsive to major economic events and structural changes. The COVID-19 pandemic’s dramatic reshaping of remote work patterns, the rise of platform and gig economy employment, the persistent gender pay gap, the minimum wage debate, and the potential displacement effects of automation and artificial intelligence have generated a rich agenda of empirically tractable dissertation questions with high policy salience. Labour economics is also one of the sub-fields with the greatest variety of available datasets — from individual-level survey data (LFS, BHPS, UKHLS) to administrative tax and wage records — making it particularly accessible for BSc and MSc dissertations.
Minimum Wage Effects on Employment: Evidence from UK Local Labour Markets
Using the 2016 introduction of the UK National Living Wage and subsequent above-inflation upratings as quasi-experiments, with variation in the “bite” of the minimum wage across local labour markets (defined as the ratio of the new minimum to the pre-existing median wage in each travel-to-work area) to estimate employment, hours, and wage effects using a DiD design following Dube, Lester, and Reich’s (2010) contiguous county approach.
Decomposing the Gender Pay Gap: The Role of Occupation, Hours, and the Child Penalty
Using UKHLS (Understanding Society) panel data to decompose the gender wage gap into explained (occupation, sector, hours, experience) and unexplained (potential discrimination) components using the Oaxaca-Blinder decomposition; testing whether the “child penalty” — the earnings cost of motherhood documented by Kleven et al. — accounts for the majority of the residual gap and whether it varies by education and occupation.
Robots, Automation, and Labour Market Polarisation: Evidence from UK Manufacturing Regions
Replicating Acemoglu and Restrepo’s (2020) robot adoption analysis for the UK manufacturing sector using International Federation of Robotics data and ONS regional employment statistics; testing whether robot adoption predicts employment polarisation (growth of high-skill and low-skill jobs with hollowing-out of middle-skill routine tasks) and whether regions with higher prior manufacturing employment are more exposed to automation-related displacement.
The Remote Work Revolution: Productivity, Career Outcomes, and Urban Wage Premiums
The COVID-19 pandemic forced an unprecedented natural experiment in remote work, generating variation in remote work adoption that is plausibly exogenous to individual productivity. This dissertation uses LFS or UKHLS data around the pandemic onset to test whether workers who shifted to remote work experienced changes in earnings growth, promotion rates, or geographical mobility — and whether the urban wage premium has narrowed as remote work enables workers to relocate away from high-rent cities while retaining urban-sector jobs.
Gig Work, Platform Employment, and Worker Welfare: Earnings Volatility and the Absence of Employment Protections
The classification of platform workers (Uber, Deliveroo, Upwork) as self-employed rather than employees has significant implications for their access to minimum wage protection, holiday pay, sick pay, and pension auto-enrolment. This dissertation uses UK Employment Tribunal records, platform earnings survey data, and the 2021 Uber Supreme Court ruling as a quasi-experiment to assess whether employment reclassification improves worker welfare — measured by earnings stability, hourly wage levels, and access to employment protections.
Health Economics Dissertation Topics: Healthcare Systems, Demand, and Policy Evaluation
Health economics applies economic tools — demand analysis, principal-agent theory, market structure analysis, and causal inference methods — to the production, financing, and consumption of healthcare. It is one of the most policy-relevant and data-rich areas of applied economics, with the COVID-19 pandemic generating a vast natural experiment in health system capacity and population health outcomes, and with NHS administrative data, US Medicare/Medicaid records, and multi-country comparative health system data providing strong empirical foundations for dissertation research at all three levels.
NHS Waiting Times and Health Outcomes: Does Queue Length Affect Mortality?
Using NHS England referral-to-treatment and mortality statistics to test whether longer waiting times for elective procedures predict worse health outcomes, with variation in waiting times across CCGs (Clinical Commissioning Groups) providing cross-sectional identification and pandemic-related waiting list surges as a temporal discontinuity.
Data: NHS England Statistics; ONS Mortality Statistics; PHE Health Profiles. Method: Panel FE with CCG-level controls; RDD around 18-week referral-to-treatment target.Pharmaceutical Pricing, Generic Entry, and Consumer Welfare in Drug Markets
How the expiry of drug patents and entry of generic manufacturers affects price competition, market structure, and consumer welfare in pharmaceutical markets; whether observed price paths after patent expiry are consistent with Bertrand competition or with branded manufacturers’ strategic responses to limit generic share.
Data: NHS BSA Drug Tariff; FDA Orange Book (US); IMS Health claims data. Method: Event study around patent expiry; price regression with generic entry dummy and count of entrants.Minimum Unit Pricing for Alcohol: Did Scotland’s 2018 Policy Reduce Harmful Consumption?
Scotland introduced minimum unit pricing (50p per unit of alcohol) in May 2018 — the first such policy in the world at national scale. Using Scotland vs. England/Wales as a DiD design with pre-policy trends as the identifying assumption, this dissertation estimates the effect on alcohol sales volumes, hospital admissions for alcohol-related conditions, and crime statistics.
Data: Public Health Scotland; NHS Scotland statistics; ONS Scottish retail sales. Method: DiD with synthetic control robustness check. A textbook natural experiment MSc topic with publicly available data.Mental Health, Productivity, and the Macroeconomic Costs of Untreated Illness
Mental health disorders are estimated to cost the UK economy over £100 billion per year in lost productivity, absenteeism, and healthcare costs. This dissertation uses individual-level UKHLS data to estimate the wage penalty associated with diagnosed mental health conditions, testing whether treatment (therapy, medication) moderates the earnings effect and estimating the aggregate output cost of the UK’s mental health treatment gap.
Data: UKHLS Understanding Society; Adult Psychiatric Morbidity Survey; IAPT treatment data. Method: Panel fixed effects with lagged health variables; matching estimator for treatment effect estimation.COVID-19 Excess Mortality and Health System Capacity: A Cross-Country Comparison
Cross-country variation in COVID-19 excess mortality (the difference between observed and expected deaths) reflects differences in health system capacity, pre-pandemic population health, government policy responses, and demographic structure. This dissertation uses WHO excess mortality estimates and OECD health system data to identify the structural health system characteristics most predictive of excess mortality outcomes across OECD and non-OECD countries.
Data: WHO Global Health Observatory; OECD Health Statistics; Our World in Data COVID-19 dataset. Method: Cross-country regression with robust SEs; Bayesian model averaging for variable selection.Health Inequality and Socioeconomic Determinants: The Social Gradient in UK Mortality
The relationship between socioeconomic position and health outcomes — the “social gradient” documented across mortality, morbidity, and self-assessed health — is one of the most robust findings in health economics. This dissertation maps the social gradient in life expectancy across UK local authority areas using PHE data, tests for widening or narrowing over the austerity period (2010–2019), and assesses whether local authority health expenditure moderates the gradient.
Data: ONS Health State Life Expectancies; PHE Profiles; HSCIC local authority expenditure. Method: Area-level regression with deprivation index controls; quantile regression for gradient heterogeneity.International Trade & Political Economy Dissertation Topics
International trade economics — studying the patterns, determinants, and welfare effects of trade flows between countries — has been revitalised by the dramatic structural changes in the global economy over the past two decades: the rise of global value chains, the China Shock, the post-2016 surge in trade protectionism, and the COVID-19 pandemic’s disruption of supply chains. Political economy — the study of how political institutions, electoral incentives, and distributional conflicts shape economic policy — intersects with trade in questions about tariff-setting, trade policy formation, and the political consequences of trade liberalisation. Both sub-fields offer rich dissertation opportunities at every academic level.
Brexit and UK Trade Flows: Evidence from Gravity Model Estimation
The United Kingdom’s exit from the European Union’s Single Market and Customs Union on 1 January 2021 provides a dramatic natural experiment in the trade effects of regulatory divergence and tariff introduction. Using a gravity model of bilateral trade flows with synthetic control methodology — constructing a counterfactual UK trade trajectory based on comparable non-Brexit countries — this dissertation estimates the causal effect of Brexit on UK goods and services exports to the EU and third-country markets, with heterogeneous effects across sectors defined by their pre-Brexit EU trade intensity and regulatory integration depth. The research contributes to the literature on the trade costs of non-tariff barriers and regulatory divergence, a question with direct relevance for ongoing UK-EU trade negotiations and for other countries considering major trade regime changes.
Supply Chain Resilience and the Deglobalisation Hypothesis: Did COVID-19 Reverse Offshoring?
The COVID-19 pandemic exposed the fragility of highly extended global supply chains — semiconductor shortages, PPE procurement failures, and food supply disruptions generated intense policy pressure toward “reshoring” and supply chain diversification. This dissertation uses World Input-Output Database data and firm-level survey evidence to test whether supply chain shortening and diversification actually occurred after 2020, or whether cost pressures have reasserted the pre-pandemic global value chain structure — with implications for the economics of trade policy and the costs of industrial policy interventions aimed at strategic supply chain security.
Trade Protection and Electoral Politics: Do Governments Raise Tariffs Before Elections?
Testing whether import tariff changes are systematically timed around election cycles in democracies, using panel data on tariff schedules and election dates across WTO member countries; the political economy model of endogenous trade policy and whether electoral incentives distort the tariff-setting process away from welfare-maximising levels toward protection of politically pivotal industries.
FDI, Technology Transfer, and Productivity Spillovers in Emerging Markets
Testing whether inward foreign direct investment generates technology spillovers that raise the productivity of domestically owned firms in the same or related industries; the horizontal vs. vertical spillover distinction; the role of absorptive capacity (human capital, R&D intensity) in determining whether FDI spillovers are positive or negative for the host economy.
Trade Liberalisation and Wage Inequality: Has Globalisation Widened the Skill Premium?
Extending the Stolper-Samuelson framework to contemporary data: testing whether trade liberalisation in middle-income countries (China, India, Brazil, Mexico) has increased the relative wages of skilled workers, using variation in industry-level import and export exposure to construct labour market shocks that can be mapped onto individual wage outcomes.
Econometric Methodology Guide: Choosing the Right Identification Strategy
Econometric methodology is not a technical add-on to an economics dissertation — it is the foundation on which every empirical claim rests. The credibility revolution in economics, associated with the work of Angrist, Pischke, Card, Krueger, and their collaborators, has established a clear hierarchy of empirical methods based on the strength of the causal identification they provide. Understanding where your research question sits in this hierarchy — and choosing a method whose assumptions you can defend — is the single most important methodological decision in any economics dissertation. The following guide maps the major identification strategies used in applied microeconomics, development economics, and policy evaluation.
Difference-in-Differences (DiD)
Causal identification from policy variation over time
DiD compares the change in outcomes for a treated group before and after treatment to the change for an untreated control group over the same period. The key identifying assumption is parallel trends — that treated and control groups would have followed the same trajectory in the absence of treatment. Modern DiD extensions (Callaway-Sant’Anna, Sun-Abraham) address staggered adoption settings where the classic two-period DiD produces biased estimates under treatment effect heterogeneity.
Example application: Estimating the effect of minimum wage increases on employment using variation in implementation timing across UK local authorities or US states.Regression Discontinuity Design (RDD)
Causal identification from threshold-based eligibility rules
RDD exploits the discontinuous assignment of treatment at a known threshold of a running variable — comparing units just above the threshold (treated) to units just below (untreated). Because units near the threshold are similar in all respects except their treatment status, the comparison provides credible local average treatment effect (LATE) estimates at the threshold. Key diagnostics: McCrary density test for sorting, balance tests on pre-determined covariates, and robustness to bandwidth choice.
Example application: Estimating the effect of pension receipt on savings behaviour using the state pension age as a discontinuity; educational returns using school entry age cutoffs.Instrumental Variables (IV)
Causal identification using exogenous shifters of endogenous variables
IV estimation uses a variable (the instrument) that is correlated with the endogenous treatment but affects the outcome only through its effect on the treatment — the exclusion restriction. Valid instruments must satisfy relevance (first-stage F-statistic > 10 as a rule of thumb), exogeneity (not directly affecting the outcome), and monotonicity (treatment moves in the same direction for all compliers). The most creative and influential IV papers find instruments that are both credible and surprising — weather, geography, historical accidents, lottery randomisations.
Example applications: Using colonial settler mortality as an instrument for institutional quality; using quarter-of-birth as an instrument for schooling; using military draft lottery as an instrument for military service.Synthetic Control Method (SCM)
Causal identification for aggregate units from comparative case studies
The synthetic control method (Abadie, Diamond, and Hainmueller, 2010) constructs a weighted combination of control units that best replicates the pre-treatment trajectory of the treated unit, then uses this synthetic counterfactual to estimate the treatment effect post-intervention. It is particularly well-suited to policy evaluations where there is a single treated unit (one country, one region, one city) and a small pool of potential controls. Inference is conducted through permutation tests (placebo in space).
Example applications: Estimating the economic impact of German reunification; evaluating the effect of California’s tobacco control programme; Brexit’s effect on UK GDP (Springford, 2020).Randomised Controlled Trials (RCTs)
Gold-standard causal identification through experimental randomisation
RCTs randomly assign treatment and control status, so that — by design — treatment is independent of all confounders, observed and unobserved. They provide the cleanest causal identification available but face external validity concerns (the effect estimated is the Local Average Treatment Effect for the study population, which may not generalise), implementation challenges in field settings, and ethical constraints in some policy domains. For dissertation purposes, RCT analysis typically uses existing experimental datasets from J-PAL, IPA, or AEA RCT Registry rather than conducting new experiments.
Example datasets: Kenya cash transfer RCTs (GiveDirectly); India microfinance trials (Banerjee et al.); Kenya deworming (Miguel and Kremer, 2004); Colombia teacher performance pay (Barrera-Osorio and Raju, 2017).Vector Autoregressions (VARs) & Time Series Methods
Dynamic modelling of macroeconomic relationships and forecasting
VAR models estimate a system of equations in which each variable is regressed on its own lagged values and the lagged values of all other variables in the system, allowing the dynamic responses of variables to exogenous shocks (impulse response functions) to be estimated. Structural VARs (SVARs) impose identifying restrictions (sign restrictions, short-run or long-run zero restrictions) to recover economically meaningful structural shocks. They are the workhorse of empirical macroeconomics for policy transmission analysis.
Example applications: Estimating monetary policy transmission using a SVAR with interest rate as policy instrument; forecasting GDP using a factor-augmented VAR (FAVAR); estimating oil price shock transmission using sign-restricted SVAR.Two Essential External Resources for Economics Dissertation Research
For literature discovery and working paper access, the NBER Working Paper Series is indispensable — it indexes over 30,000 working papers across all economics sub-fields, providing access to frontier research often months or years before journal publication, and is the primary vehicle through which cutting-edge empirical economics is first disseminated to the profession. For data and country-level statistics, the World Bank Open Data portal provides free access to over 1,600 indicators covering economic, social, environmental, and institutional characteristics for over 200 countries from 1960 to the present — making it the single most useful starting data source for cross-country development, macroeconomics, and trade dissertations at all three levels.
A Five-Chapter Economics Dissertation Structure
Motivate the research question with a specific empirical puzzle or policy problem. State your research question precisely — naming the outcome, treatment, population, and identification strategy. Summarise your main findings and state their contribution to the existing literature. Do not provide background context without connecting it to why your specific question matters.
Map the relevant theoretical and empirical literature. Organise thematically around the key debates, not source by source. Identify precisely where your contribution fits — what question you are answering that the existing literature has not, or what new context or identification strategy you bring to a question already studied elsewhere.
Describe your data sources, sample construction, variable definitions, and summary statistics. Specify your econometric model and identification strategy, justify the key assumptions, and describe how you will test their validity. Present your first-stage results (for IV), density tests (for RDD), or parallel trends evidence (for DiD) here.
Present main results, robustness checks (alternative specifications, sample restrictions, placebo tests), and heterogeneity analysis. Interpret magnitudes alongside statistical significance — economic significance (effect size relative to the sample mean) matters more than p-values alone. Present tables and figures that communicate results clearly without requiring the reader to reverse-engineer the econometric specification.
Summarise your main empirical findings without repeating all the detail from the results chapter. Discuss the policy implications explicitly — what should decision-makers do differently given your findings? Acknowledge limitations honestly, including potential threats to identification that robustness checks could not fully resolve. Identify the most productive directions for future research your results open.
Economics Dissertation Thesis Statement Templates: From Research Area to Defensible Claim
A strong economics dissertation research question is not a broad topic announcement — it is a precise, answerable empirical or theoretical question that specifies the outcome of interest, the treatment or independent variable, the population, the time period, and the identification strategy. The following templates demonstrate the difference between vague topic formulations and precise research questions across the major economic sub-fields and dissertation levels.
Economics Dissertation Research Question Templates — Strong vs. Weak
Precise research questions and vague topic formulations across sub-fields and dissertation levels
Labour Economics — Minimum Wage (BSc)
BSc LevelEnvironmental Economics — Carbon Pricing (MSc)
MSc LevelDevelopment Economics — Institutions (PhD)
PhD LevelBehavioural Economics — Nudge Design (BSc/MSc)
BSc/MScSeven Critical Mistakes in Economics Dissertations — and How to Fix Each One
| # | ❌ Mistake | Why It Costs Marks | ✓ The Fix |
|---|---|---|---|
| 1 | Confusing correlation with causation in empirical results | The most frequent and most serious error in BSc and MSc economics dissertations is presenting OLS regression results as evidence of causal effects when no credible identification strategy has been established. Running a regression of GDP growth on trade openness, reporting R² = 0.62, and concluding that “trade drives growth” confuses a correlation with a causal claim — a mistake that every econometrics course explicitly warns against but that still appears in a majority of undergraduate dissertations. Examiners know the difference, and dissertations that make causal claims without identification lose marks at every mark scheme level. | Either adopt a credible identification strategy (DiD, RDD, IV, or RCT) that provides genuine causal identification, or be explicit and consistent about the descriptive or correlational nature of your findings. “This result is consistent with the hypothesis that X causes Y, but we cannot rule out reverse causality or omitted variable bias without a credible instrument for X” is a valid and honest statement. “X causes Y” without identification is not. |
| 2 | Choosing a topic with no accessible data | Selecting a dissertation topic on the basis of intellectual interest alone — without verifying that the required data actually exists and is accessible within the project’s timeline and budget — leads to topics that are abandoned mid-project, dramatically rescoped under time pressure, or completed with wholly inadequate empirical foundations. This is the single most avoidable cause of poor economics dissertations at every level. | Before committing to any topic, spend two hours searching the data sources relevant to your question: World Bank Open Data, ONS, UK Data Service, FRED, J-PAL Dataverse, EconLit. If you cannot identify a specific dataset, a specific sample, and a specific set of variables that can answer your research question, the topic is not viable. Change the question, not the methodology, until you find a combination that works. |
| 3 | Treating statistical significance as economic significance | A coefficient that is statistically significant at the 1% level (very unlikely to be zero) is not necessarily economically significant (meaningful in magnitude). An estimated effect of minimum wage increases that reduces employment by 0.001 percentage points with a t-statistic of 4.2 is statistically significant but economically negligible. Dissertations that obsess over stars (*, **, ***) while ignoring effect sizes, confidence intervals, and substantive interpretation of magnitudes misunderstand what econometric results mean. | For every coefficient you report, contextualise its magnitude: express it as a percentage of the sample mean, compare it to effects of similar policies in the literature, or calculate what it implies in real-world terms (X additional people employed, £Y in additional household income per year). A well-interpreted estimate with a wide confidence interval is worth more than a precisely estimated effect whose size is never explained. |
| 4 | Presenting a literature review as an annotated bibliography | A literature review that describes what each paper finds — “Smith (2019) found that X. Jones (2020) found that Y. Brown (2021) also examined X and found Z” — demonstrates reading but not synthesis or critical engagement. Economics examiners want to see that you understand the structure of the literature: the key debates, the evolution of the empirical methodology, the contested findings, and precisely where your contribution fits. A source-by-source summary provides none of this. | Organise your literature review thematically around debates and methodological evolution: “The early literature estimated the effect of X using OLS with cross-country data, finding consistently positive correlations (Smith, 2001; Jones, 2003; Brown, 2005). More recent work has challenged the causal interpretation of these estimates, exploiting natural experiments to show that the OLS estimates are substantially upward biased due to reverse causality (Garcia, 2018; Lee, 2020). My contribution sits within this second wave, extending Garcia’s (2018) identification strategy to a new institutional context.” |
| 5 | Ignoring the parallel trends assumption in DiD designs | Difference-in-differences rests entirely on the parallel trends assumption — that treated and control groups would have followed the same trajectory in the absence of treatment. Presenting DiD results without testing or even acknowledging this assumption is a methodological gap that examiners will always identify. The assumption is untestable in the post-treatment period, but it can be evaluated using pre-treatment data — and failing to do so leaves the entire causal interpretation undefended. | Always present pre-treatment trend plots and formal tests of parallel trends as part of your DiD analysis. Use event-study specifications (plotting the estimated treatment effect at each relative time period before and after the policy) to show that there were no significant pre-trends before the treatment date. If pre-trends are present, discuss what they imply for the interpretation and consider alternative estimation approaches (synthetic DiD, augmented synthetic control) that are more robust to pre-trend violations. |
| 6 | Selecting instruments without satisfying the exclusion restriction | IV estimation produces consistent estimates only if the instrument affects the outcome exclusively through its effect on the treatment variable — the exclusion restriction. This assumption is inherently untestable and must be defended on theoretical grounds. Dissertations that select instruments because they are correlated with the endogenous variable, without carefully examining all the ways the instrument might affect the outcome directly, produce IV estimates that may be more biased than OLS. | For every IV specification, explicitly articulate all the pathways through which the instrument might affect the outcome other than through the treatment, and explain why each of these is either absent or negligible. Consider sensitivity analysis using bounds approaches (Conley et al. relaxing the exclusion restriction) or over-identification tests where multiple instruments are available. A well-defended weak exclusion restriction is better than an unexamined strong one. |
| 7 | Failing to discuss policy implications of empirical findings | Economics is, ultimately, a social science — and the purpose of empirical economics research is not merely to produce estimated coefficients but to inform decisions about economic institutions and policy. Dissertations that present results technically but never translate them into policy-relevant statements miss the most important audience for economics research and fail to demonstrate the ability to connect empirical findings to the real-world decisions they should inform. | Every results section and conclusion should include a paragraph that asks: what do these findings imply for the policy question that motivated the research? Are the effects large enough to justify the policy costs? Does the heterogeneity analysis suggest that the policy should be targeted differently? Are there unintended consequences visible in the data that policymakers should be aware of? Economic findings without policy implications are incomplete economic arguments. |
Pre-Submission Economics Dissertation Checklist
- Research question specifies outcome, treatment, population, and identification
- Data sources identified, accessed, and adequately described in data section
- Literature review synthesises debates, not just describes papers
- Identification strategy assumptions stated and tested where possible
- Causal claims made only where identification supports them
- Effect sizes interpreted in economic terms, not just statistically
- Robustness checks presented (alternative specs, sample restrictions, placebos)
- For DiD: parallel trends plot and event-study included
- For IV: first-stage F-statistic reported; exclusion restriction defended
- Summary statistics table presented before results
- Policy implications discussed explicitly in conclusion
- Limitations of identification strategy acknowledged honestly
FAQs: Economics Dissertation Questions Answered
Conclusion: Your Economics Dissertation as Original Economic Thinking
The economics dissertation — at any level from BSc to PhD — is the assessment in which you move from consuming economics to producing it. It demands something that problem sets, exams, and essays do not: a genuine research question you have formulated yourself, a dataset you have assembled and cleaned yourself, and an empirical or analytical strategy you have chosen, implemented, and defended yourself. That process is difficult, time-consuming, and often frustrating — but it is also uniquely rewarding, because it is the first time most economics students experience what it actually feels like to make an original contribution to economic understanding, however modest.
The 100+ topics, methodology frameworks, research question templates, and strategic guidance in this guide are not a substitute for that work — they are a starting map. The actual journey requires you to read deeply in one sub-field, understand one empirical literature, find one dataset, implement one identification strategy, and draw one set of conclusions with the intellectual honesty that characterises good empirical economics. The quality of that journey — not the breadth of your topic or the sophistication of your software package — determines the quality of your dissertation. Every excellent economics dissertation is excellent in the same way: it asks a specific, important question; answers it with the best available data and the most credible identification strategy the data allows; and interprets the results with precision, honesty, and genuine economic insight.
For expert support with every stage of your economics dissertation — from research question development and literature review through data analysis, econometric specification, and final editing — the specialist team at Smart Academic Writing is ready to help. Explore our dissertation writing service, thesis coaching, data analysis support, literature review help, and economics homework help today.