Introduction to Public Economics Lecture 1 PDF

Today’s agenda Course information and administration Theoretical framework Introduction to public economics First theorem of welfare economics Two-good pure exchange (endowment) economy Social welfare function New trends in public economics Evidence-based policy mak...

Today’s agenda Course information and administration Theoretical framework Introduction to public economics First theorem of welfare economics Two-good pure exchange (endowment) economy Social welfare function New trends in public economics Evidence-based policy making “Credibility revolution”  What is public economics? What does the public mean? Community I = {1, 2, …, N}: At least two players, N ≥ 2. What is economics? Resource allocation (Q) Commodity = {wheat, meat} Community = {John, Kate} Q = Commodity × Community = {wheat, meat} × {John, Kate} Wheat (X) Meat (Y) John (A) XA YA Kate (B) XB YB  What is public economics? What is public economics? Policy τ: Involves policies to improve the welfare through resource allocation Tools of τ include tax (fees, charges, contribution, subsidy) market or industry regulation To achieve desirable allocation Q(τ) Price system would be affected P(τ) Who makes the policy? Usually involve a additional player: government Put aside the self-organized public policy SAFETY INSTRUCTION TO STUDENTS IN UWA BUSINESS SCHOOL Theoretical Framework  Why public economics? Starting point (Benchmark): The first theorem of welfare economics Original idea: Adam Smith’s “invisible hand” Desirable resource allocation Q(P) can be automatically achieved through the market equilibrium price P (“invisible hand”) Q = Commodity × Community = {wheat, meat} × {John, Kate} P = prices of Commodity = {price of wheat, price of meat} No room for public economics unless the “invisible hand” fails (market failure) Three fundamental questions in public economics Why may free market not work well and extra policy τ need to be introduced? Market failure: public goods, externality, asymmetric information, monopoly What is the effect of the policy τ on the market economy (Q, P)? Tax, social security, expenditure How to design an optimal policy τ? More advanced questions Why is the optimal policy not proposed or implemented? Political economy: politicians and institutions Development economics: special economic constraints in developing countries  First Theorem of Welfare Economics Why learn the First Theorem? General analytical framework: simple, abstract theory for reasoning Benchmark for studying market failure and government polices First Theorem of Welfare Economics: What does it mean? If there are markets for all goods and all markets are perfectly competitive, then an economy in equilibrium is in a Pareto-efficient state. Pareto efficient A situation where no one can be made better off without making someone else worse off. Also known as ‘Pareto optimal’ or ‘the Pareto criterion’ Pareto improvement A relative statement indicating that an alternative situation is superior to the current situation if everyone is better off in the alternative situation.  Two-good economy Two-good pure exchange (endowment) economy Extension (ignored in this lecture) Economy with production factor(s) The economy (2×2 economy) Community: 2 consumers = {Consumer A, Consumer B} Commodity: 2 goods = {Wheat (X) and Meat (Y)} Endowment (Initial resource allocation) Wheat (X) Meat (Y) Consumer A X 0A Y0A Consumer B X 0B Y0B  Pause! Initial Allocation of Chocolate (Random)  Each consumer’s problem -- Graph Optimal condition Tangent point between Meat Y U0 Indifference curve U1 the indifference curve and the budget line Q0 Y0A MRS XA ,Y  ERS X ,Y Note: MRS depends on Q1 Y1A the preference of the consumer the consumption choice ERS is independent of the preference of the consumer Consumer A X 0A X 1A Wheat X the consumption choice  Each consumer’s problem -- Algebra Each consumer i is to choose a most desirable consumption bundle ( X 1i , Y1i ) , i  { A, B} Subject to Preference: utility function ui  U i ( X i ,Y i ) , i  { A, B} Endowment: ( X 0i , Y0i ) , i  { A, B} Market price: ( PX , PY )  Each consumer’s problem Maximize utility ui  U i ( X i ,Y i ) , i  { A, B} Subject to budget constraint (market price and endowment) PX X i  PY Y i  PX X 0i  PY Y0i  PX  X i  X 0i   PY Y0i  Y i  expenditure market value of endowment expenditure on buying X revenue from selling Y Optimal choice (resource allocation) if X i  X 0i U i ( X i , Y i ) / Y i PX  MRS Xi ,Y  ERS X ,Y  U ( X , Y ) / X i i i i PY  Edgeworth box Put two consumers together and let them trade! How much each consumer finally consumes (allocation)? The total endowment of this economy is given as ( X 0 , Y0 ) :  X 0A  X 0B  X0  Y  0 A  Y0B  Y0 Wheat (X) Meat (Y) Consumer A X 0A Y0A Consumer B X 0B Y0B Total Endowment X0 Y0  Edgeworth box x0B Good X Dealer B Good Y y0B Initial Allocation α0 y0A  y0B y0A Good Y DealerA Good X A x0 xA  xB 0 0 Acknowledgements: Thanks to Prof. Michael McLure for sharing the graphs of Edgeworth box.  Edgeworth box The essence of the Edgeworth box Put an 2×2 economy in a box! The box can represents any resource allocation and trade in the economy We can place any allocation (α) in an ‘Edgeworth box’ diagram where α = αA + αB This is achieved by setting Corner South West Corner as the origin of the allocation plane for dealer A North East Corner as the origin of the allocation plane for dealer B Border The borders of the box as the coordinates of the overlapped X-Y planes of dealers A and B.  The Edgeworth Box – Problem for DealerA x0B Good X Dealer B y0B Good Y α0 U 5A ly A 0 U 4A U 3A Good Y U1A U 2A DealerA Good X x0A  The Edgeworth Box – Problem for Dealer B x0B Good X Dealer B y0B Good Y α0 U1B ly A 0 U 2B U3B Good Y U 5B U 4B DealerA Good X x0A  The Edgeworth Box – Pareto Improvements x0B Good X Dealer B U1B y0B Good Y U B 2 α0 Pareto B Improvements U 5A U 5 ly A i.e. Potential 0 Welfare gains from trade U 2A Good Y U1A DealerA Good X x0A Edgeworth Box – Pareto Improvements & the Contract Curve x0B Good X Dealer B y0B Good Y α0 Contract curve All tangent points between two ly A indifference curves. 0 MRS XA ,Y  MRS XB ,Y Good Y DealerA Good X A x 0  The Edgeworth Box – Contract Curve Good X Dealer B Good Y Contract Curve Good Y DealerA Good X  Optimal consumption Optimal consumption choice MRS between two goods = ERS (= relative price) PX MRS i X ,Y  ERS i X ,Y  PY Relative price Wheat (X) Meat (Y) between X and Y  PX Consumer A MRS XA ,Y PY || ||  PX B Consumer B MRS X ,Y PY  The First Theorem Contract curve = {allocation that is Pareto optimal} Every point along that curve within the ‘lozenge’ around the initial allocation is a Pareto improvement Any movement off the contract curve will necessarily harm one person Optimal consumption choice given the market price PX MRS Xi ,Y  PY Market (price signal) is a great coordinator between consumers 1St Theorem: Market equilibrium is Pareto optimal The point on the contract curve that can be obtained by trade is the point where the price ray, which passes through the initial allocation, intersects the contract curve PX MRS A X ,Y   MRS XB ,Y PY  Edgeworth Box – Pareto Improvements Through Trade △x B Good X Dealer B Good Y U 3B Market and Pareto α0 allocation B U △y A 4 △y B Market equilibrium is Pareto optimal α1 PX MRS XA ,Y   MRS XB ,Y PY U 4A A U 3 Good Y DealerA Good X △x A Price ray  First Theorem of Welfare Economics What does the Pareto efficiency mean? No “win-win” or Pareto improvement situation any more Only one of the efficiency criteria Another criterion: Karldo-Hicks improvement Less stringent requirement: allow for “win-lose” with win > lose, so that the loss may be compensated from the gain Lack of equity consideration No “win-win” includes the situation where a winner takes all! Can we have both efficiency and equity? Market may not be sufficient Requires optimal policy design Social welfare function to combine both objectives From Contract Curve to Utility Possibility Frontier Good X Dealer B Good Y F C3 C2 E C1 Contract Curve Good Y DealerA Good X  Utility possibility frontier UB C1 F C2 E C3 UA  Social welfare function UB Indifference curve of social welfare W1 W0 Social Welfare Function W  F (u A , u B ) UA  Social welfare function (SWF) SWF differs according to their answers to the following question Can rich people’s utility substitutable to that of the poor? Completely yes Utilitarian SWF W  u A  uB Absolutely no Rawlsian SWF W  min{u A , u B } Partly yes Atkinson SWF (mixed)  1 W  u  A 1  u  B 1 1 ,  0  Social optimal UB Indifference curve of social welfare W1 W0 C1 F C2 C3 E UA New Trends  Evidence-based policy making What is evidence-based policy (EBP)? Policy decisions should be based on rigorously established objective evidence In contrast to policymaking based on ideology or 'common sense' Attempt to fill the longstanding gap between Policy making mainly driven by political factors Policy suggestions based on theories and reasoning Australia is a pioneer EBP (Productivity Commission, 2010)  Evidence-based policy making Why is EBP emerging? – Academic perspective and background “Credibility revolution” in empirical economics (Angrist and Pischke, 2010) Center of the revolution Transparent Research Design like in natural science Research question should be clearly defined Causal effect of X on Y (e.g.: X = corporate income tax, Y = firm investment) Good measurements of X and Y Identification strategy Look for random variation in X and compare the outcome in Y Random: No confounder Z that may also affect Y along with X (so X is correlated with Z, e.g.: Z = interest rate)  Evidence-based policy making Conditions of EBP and Research trends More data Survey data Administrative data Big data: Merge data from various sources More transparent method Identification: More transparent empirical method on causal effects of policies Experiments: lab experiment, field experiment, natural experiment Machine learning: data driven instead of model driven Extended theory basis More behavior economics Warning: traditional theory is still powerful and suggestive Broader Topics (mainly due to data availability) Broader and more balanced policy evaluation More integration of public economics and development economics  What Are We Talking About In Public Economics? (Kleven, 2018) Understanding “new directions in research” based on word and language trends Textual analysis of public economics papers since 1975 NBER (National Bureau of Economic Research) working papers Focus on PE(Public Economics)-tagged papers (4676 papers) Analyse full texts Caveat NBER is a selected sample of the profession, and the nature of the selection has changed over time We Talk Less about Taxes than We Used to Do When We Do Talk About Taxes, Which Taxes Do We Talk About? We Talk Mostly about the US, But Less So over Time When We Don't Talk about the US, Who Do We Talk about?  The Identification Revolution Note: The graph shows the fraction of papers that mention the word "identification" in the context of empirical identification. The Rise of Experiments The Rise of Quasi-Experiments The Rise of Administrative Data Big Data & Machine Learning The Rise of Behavioural Economics  Identification The idea of randomized experiment in natural sciences To identify the treatment effect with two randomized groups A treatment group A control group Randomized The two groups are statistically similar Treatment effect The difference between two groups after the treatment (e.g. taking a new vaccine) Methods (Duflo, Glennerster, and Kremer, 2008) RCT: Randomized control trial DID: Difference-in-Difference RDD: Regression discontinuity design IV: Instrument variable  Machine learning (ML) Traditional empirical study Model driven OLS: Y = α + β X + ε Uncertainty in parameter: OLS: estimate β Machine learning – More in Week 9 Data driven Observe {Yi , Xi } Uncertainty in model Yi = f (Xi, θ) Even the function form f (Xi, θ) is unknown (e.g.: digit recognition with neural network)  EBP + ML Evidence-based policy making + machine learning Androutsopoulou, Aggeliki, Yannis Charalabidis (2018)  Summary Theoretical framework Introduction to public economics Resource allocation: Community × Commodity Public policy τ to achieve desirable allocation when market fails or allocation is not equal First theorem of welfare economics 2×2 economy Pareto optimal allocation is achieved through decentralized market economy Common market signal coordinates all private heterogeneous economic agents Social welfare function and social optimality To deal with both efficiency and equity Theoretical framework for policy design New trends in public economics EBP: Evidence-based policy making “Credibility revolution” EBP + ML