1. Consumption, Production, Welfare B: Consumer Behaviour Univ. Prof. dr. Maarten Janssen University of Vienna Winter semester 2013

2. Consumer behaviour: Overview Often it is important to know how consumers react: – Does raising taxes on cigarettes, alcohol increase government revenues, and if so by how much? – Is it effective to reduce gasoline consumption to raise gasoline tax and refund consumers for the higher prices? If we want to assist young families with newborn babies, it is more effective to give them a lump sum amount or to provide them with a week of nursing care Is it appropriate to interpret area under demand curve as a measure of consumer welfare? Most of these questions require to go beyond the demand curve and inquire into consumer behaviour more deeply

3. Consumer behaviour: theory Constraint optimization – Constraints: usually budget, but other constraints also possible (in policy examples may be important) – Optimization: what does consumer want? Notion of preference (well-being; Bentham) is important Preference relationship > – x>y : x is strictly preferred to y – x≥y : x is weakly preferred to y – x~y : consumers is indifferent between x and y

4. Axioms on preference relation Completeness: for all possible x,y x≥y or y≥x Transitivity: for all possible x,y,z if x≥y and y≥z, then x≥z Are these axioms realistic? What would constitute violations? Why are these axioms made?: – Without them, a notion of consumer well-being is difficult and welfare questions hard to address – Formally, if a preference relation does not satisfy these axioms, they cannot be described by a utility function – Transitivity is strongly connected to rationality (violation and the money pump argument)

5. What is a utility function? Way to describe a preference relation using the mathematically convenient way of functions (so we can use the corresponding mathematical operations). – Utility function u(.) describes preference relation ≥ if u(x) ≥ u(y) if, and only if, x≥y – Any monotomic transformation of u(.) describes the same preference relation Can any rational preference relation be described by a utility function? – No, lexicographic preferences (essentially two- dimensional) Utility function is often graphically represented by a set of indifference curves

6. Other frequently used assumptions regarding preferences Local non-satiation: for any choice option, there is an alternative choice that if preferred Monotonicity: If x» y, then x>y Convexity: If x≥y and z≥y, then α x +(1- α) y ≥z for 0 < α < 1. As said some notion of preferences is needed to be able to evaluate policy questions

7. Consumer Theory without preferences Revealed preference: We can look at choices made of individuals and ask whether they satisfy some natural consistency requirements General: if in two choice situations, x and y were both included, and in one choice situation the agent chose x, then he cannot uniquely choose y in the other situation Under a budget constraint: if at prices p and wealth level w, individual chose x(p,w) and if at prices p’ and wealth level w’ individual chose x(p’,w’), then px(p’,w’) ≤ w implies p’x(p,w) ≤ w’ Graphical illustration budget constraint two goods

8. Implications for demand theory Does RP imply that demand curves are downward sloping? Graphical illustration Only if price changes are compensated by wealth changes Slutsky compensation: you compensate agent so that she can just afford old consumption bundle, i.e., p’x(p,w) = w’ RP in this case implies (p’ – p)[x(p’,w’) - x(p,w)] ≤ 0

9. Recent study by Wieland Müller et al. (AER 2013, forthcoming) Internet experiments with large sample of Dutch population, of which researchers know many features (age, education, wealth, etc.) They perform RP tests in choice situations Who is more rational (is more consistent with RP)? – Younger, more educated people Doing well in RP tests correlates well with wealth of individuals (if corrected for age, education and other features)

10. Implication Equivalence RP has empirical implications (that can be violated) Utility maximization under a budget constraint has empirical implications These empirical implications are almost identical

11. Maximization implication

12. Application: gasoline tax proposal under president Carter Carter proposed to increase gasoline tax to reduce use of gasoline in USA Critique: the poor can then not afford to have a car (as they cannot afford to pay gigher gasoline price) Carter reacted by saying that the poor will be income compensated for the tax increase Critique’s then said that the whole proposal is then ridiculous as it is ineffective: if people can afford the same amount as before, they will. Who is right?

13. Econ Questions and Analysis Gasoline consumption Other goods Original budget line 1.Will the consumption of gasoline decrease after an gasoline price tax? 2.If consumers are compensated will they consume less gasoline? - How are they compensated? 3.If they are Slutsky compensated, will they consume less? 4.How to reconclide answers to 1 and 3? 5.If they are Slutsky compensated, will government run a deficit over this policy?