Budget constraint, preferences, utility Varian, Intermediate Microeconomics, 8e, chapters 2, 3, and 4 1 / 43 In this lecture, you will learn • what budget set and budget line are • how their shape is influenced by taxes and food stamps • what preferences are and how they are derived • what the basic types of preferences are – why some indiference curves are straight and some curved, or circle-shaped • what we need a utility function for • how to find out whether to reconstruct a stadium 2 / 43 Budget constraint We assume that the consumer chooses a bundle (x1, x2), where x1 and x2 are quantities of goods 1 and 2. Budget constraint is p1x1 + p2x2 ≤ m: • p1 and p2 are prices of goods 1 and 2 • m is income Budget set – bundles for which: p1x1 + p2x2 ≤ m. Budget line (BL) – bundles for which: p1x1 + p2x2 = m. 3 / 43 Budget set and budget line (graph) Budget line: p1x1 + p2x2 = m 4 / 43 Budget set and budget line (graph) Budget line: p1x1 + p2x2 = m ⇐⇒ x2 = m/p2 − (p1/p2)x1 4 / 43 Composite good The theory works for more than two goods. How to plot it in a 2D graph? 5 / 43 Composite good The theory works for more than two goods. How to plot it in a 2D graph? On the y axis we can plot the composite good = money value of all other consumed goods. 5 / 43 Change in income A rise in income from m to m 6 / 43 Change in income A rise in income from m to m =⇒ parallel shift out 6 / 43 Change in price A rise in price from p1 to p1 7 / 43 Change in price A rise in price from p1 to p1 =⇒ pivot around the vertical intercept 7 / 43 Change in more variables Multiplying all prices and income by t... tp1x1 + tp2x2 = tm 8 / 43 Change in more variables Multiplying all prices and income by t does not change BL: tp1x1 + tp2x2 = tm ⇐⇒ p1x1 + p2x2 = m 8 / 43 Change in more variables Multiplying all prices and income by t does not change BL: tp1x1 + tp2x2 = tm ⇐⇒ p1x1 + p2x2 = m Multiplying all prices by t... tp1x1 + tp2x2 = m 8 / 43 Change in more variables Multiplying all prices and income by t does not change BL: tp1x1 + tp2x2 = tm ⇐⇒ p1x1 + p2x2 = m Multiplying all prices by t has the same effect as dividing income by t: tp1x1 + tp2x2 = m ⇐⇒ p1x1 + p2x2 = m t 8 / 43 Numeraire Any price or income can be normalized to 1 and adjust all variables so that the BL stays the same. Numeraire = an item with its value normalized to 1 9 / 43 Numeraire Any price or income can be normalized to 1 and adjust all variables so that the BL stays the same. Numeraire = an item with its value normalized to 1 Budget line p1x1 + p2x2 = m: • Good 1 is numeraire – the same BL: x1 + p2 p1 x2 = m p1 • Good 2 is numeraire – the same BL: p1 p2 x1 + x2 = m p2 • The income is numeraire – the same BL: p1 m x1 + p2 m x2 = 1 9 / 43 Taxes and subsidies Three types of taxes: • Quantity tax – consumer pays the amount t for each unit. → Price of good 1 increases to p1 + t. 10 / 43 Taxes and subsidies Three types of taxes: • Quantity tax – consumer pays the amount t for each unit. → Price of good 1 increases to p1 + t. • Value tax (ad valorem) – consumer pays a share τ of price. → Price of good 1 increases to p1 + τp1 = (1 + τ)p1. 10 / 43 Taxes and subsidies Three types of taxes: • Quantity tax – consumer pays the amount t for each unit. → Price of good 1 increases to p1 + t. • Value tax (ad valorem) – consumer pays a share τ of price. → Price of good 1 increases to p1 + τp1 = (1 + τ)p1. • Lump-sum tax – the value of the tax is independent from consumer’s choice. → Consumer income decreases by the size of the tax. 10 / 43 Taxes and subsidies Three types of taxes: • Quantity tax – consumer pays the amount t for each unit. → Price of good 1 increases to p1 + t. • Value tax (ad valorem) – consumer pays a share τ of price. → Price of good 1 increases to p1 + τp1 = (1 + τ)p1. • Lump-sum tax – the value of the tax is independent from consumer’s choice. → Consumer income decreases by the size of the tax. Subsidy = a tax with a negative sign 10 / 43 Rationing If there is rationing imposed on good 1, no consumer is allowed to buy a higher quantity of good 1 than ¯x1. 11 / 43 Rationing If there is rationing imposed on good 1, no consumer is allowed to buy a higher quantity of good 1 than ¯x1. 11 / 43 Taxing consumption greater than ¯x1 If consumer pays a tax only on the consumption of good 1 that is in excess of ¯x1... 12 / 43 Taxing consumption greater than ¯x1 If consumer pays a tax only on the consumption of good 1 that is in excess of ¯x1, budget line is steeper to the right of ¯x1. 12 / 43 CASE: The food stamp program Before 1979 (left graph): • value subsidy – people pay a part of the value of the food stamp • rationing – maximum value of stamps (e.g. 153 $) 13 / 43 CASE: The food stamp program Before 1979 (left graph): • value subsidy – people pay a part of the value of the food stamp • rationing – maximum value of stamps (e.g. 153 $) 13 / 43 CASE: The food stamp program Before 1979 (left graph): • value subsidy – people pay a part of the value of the food stamp • rationing – maximum value of stamps (e.g. 153 $) 13 / 43 CASE: The food stamp program Before 1979 (left graph): • value subsidy – people pay a part of the value of the food stamp • rationing – maximum value of stamps (e.g. 153 $) After 1979 (right graph) – a specific number of food stamps for free 13 / 43 CASE: The food stamp program Before 1979 (left graph): • value subsidy – people pay a part of the value of the food stamp • rationing – maximum value of stamps (e.g. 153 $) After 1979 (right graph) – a specific number of food stamps for free 13 / 43 Preferences Consumers compare bundles according to their preferences. Preference relations – three symbols: • bundle X is strictly preferred to bundle Y : (x1, x2) (y1, y2) • bundle X is weakly preferred to bundle Y (bundle X is at least as good as bundle Y ): (x1, x2) (y1, y2) • consumer is indiferent between bundles X and Y : (x1, x2) ∼ (y1, y2) 14 / 43 Assumptions about preferences Assumptions that allow ordering of bundles according to preferences: • Completeness — any two bundles can be compared: (x1, x2) (y1, y2), or (x1, x2) (y1, y2), or both 15 / 43 Assumptions about preferences Assumptions that allow ordering of bundles according to preferences: • Completeness — any two bundles can be compared: (x1, x2) (y1, y2), or (x1, x2) (y1, y2), or both • Reflexivity — each bundle is at least as good itself: (x1, x2) (x1, x2) 15 / 43 Assumptions about preferences Assumptions that allow ordering of bundles according to preferences: • Completeness — any two bundles can be compared: (x1, x2) (y1, y2), or (x1, x2) (y1, y2), or both • Reflexivity — each bundle is at least as good itself: (x1, x2) (x1, x2) • Transitivity — if (x1, x2) (y1, y2) and (y1, y2) (z1, z2), then (x1, x2) (z1, z2) 15 / 43 Weakly preferred set and indifference curves 16 / 43 Two indifference curves cannot cross Two different IC such that X Y . Why cannot they cross? 17 / 43 Two indifference curves cannot cross Two different IC such that X Y . Why cannot they cross? It follows from transitivity that if X ∼ Z and Z ∼ Y then X ∼ Y . 17 / 43 Examples of preferences – perfect substitutes Willingness to substitute one good for the other at a constant rate 18 / 43 Examples of preferences – perfect substitutes Willingness to substitute one good for the other at a constant rate =⇒ constant slope of the indifference curve (not necessarily −1). 18 / 43 Examples of preferences – perfect complements Consumption in fixed proportions (not necessarily 1:1). 19 / 43 Examples of preferences – perfect complements Consumption in fixed proportions (not necessarily 1:1). 19 / 43 Examples of preferences – bads The consumer likes pepperoni but does not like anchovies, they are a bad for her. 20 / 43 Examples of preferences – bads The consumer likes pepperoni but does not like anchovies, they are a bad for her. 20 / 43 Examples of preferences – neutrals The consumer likes pepperoni but is neutral about anchovies, they are a neutral for her. 21 / 43 Examples of preferences – neutrals The consumer likes pepperoni but is neutral about anchovies, they are a neutral for her. 21 / 43 Examples of preferences – satiation point Satiation point is the most preferred point (¯x1, ¯x2). When the consumer has too much of one of the goods, it becomes a bad. 22 / 43 Examples of preferences – satiation point Satiation point is the most preferred point (¯x1, ¯x2). When the consumer has too much of one of the goods, it becomes a bad. 22 / 43 Examples of preferences – discrete goods A discrete good is not divisible – consumption in integer amounts: • indiference ” curves“ 23 / 43 Examples of preferences – discrete goods A discrete good is not divisible – consumption in integer amounts: • indiference ” curves“ – a set of discrete points 23 / 43 Examples of preferences – discrete goods A discrete good is not divisible – consumption in integer amounts: • indiference ” curves“ – a set of discrete points • a weakly preferred set 23 / 43 Examples of preferences – discrete goods A discrete good is not divisible – consumption in integer amounts: • indiference ” curves“ – a set of discrete points • a weakly preferred set – a set of line segments 23 / 43 Well-behaved preferences Assumptions of well-behaved preferences: monotonicity and convexity 24 / 43 Well-behaved preferences Assumptions of well-behaved preferences: monotonicity and convexity Monotonicity – more is better (it excludes bads) =⇒ indifference curves have negative slope. 24 / 43 Well-behaved preferences (cont’d) Convexity – if (x1, x2) ∼ (y1, y2), then it holds for all 0 ≤ t ≤ 1 that (tx1 + (1 − t)y1, tx2 + (1 − t)y2) (x1, x2). 25 / 43 Well-behaved preferences (cont’d) Convexity – if (x1, x2) ∼ (y1, y2), then it holds for all 0 ≤ t ≤ 1 that (tx1 + (1 − t)y1, tx2 + (1 − t)y2) (x1, x2). Strict convexity – if (x1, x2) ∼ (y1, y2), then it holds for all 0 ≤ t ≤ 1 that (tx1 + (1 − t)y1, tx2 + (1 − t)y2) (x1, x2). 25 / 43 Marginal rate of substitution Marginal rate of substitution (MRS) = slope of the indifference curve: MRS = ∆x2 ∆x1 = dx2 dx1 Diminishing marginal rate of substitution – absolute value of MRS decreases as we increase x1. 26 / 43 Interpretation of marginal rate of substitution Interpretation of MRS: • The amount of good 2 one is willing to pay for one unit of good 1. • If good 2 is measured in money: MRS = marginal willingness to pay = how many dollars you would just be willing to give up for an additional unit of good 1. 27 / 43 APPLICATION: Build a stadium for Minnesota Vikings? The club does not like the stadium – considers leaving Minnesota. Fenn a Crooker (SEJ, 2009) measure how much households are willing to pay for Vikings staying in Minnesota = MRS between composite good and Vikings in Minnesota. MRS of an average household: 531 $ Value of the stadium: 531 $ × 1,323 million households = 702 mil. $ 28 / 43 APPLICATION: Build a stadium for Minnesota Vikings? The club does not like the stadium – considers leaving Minnesota. Fenn a Crooker (SEJ, 2009) measure how much households are willing to pay for Vikings staying in Minnesota = MRS between composite good and Vikings in Minnesota. MRS of an average household: 531 $ Value of the stadium: 531 $ × 1,323 million households = 702 mil. $ Estimated costs are 1 billion $. The new stadium opens in 2016 – the state provided 500 million $. 28 / 43 Utility Two concepts of utility: Cardinal utility – attach a significance to the magnitude of utility: • difficult to assign the magnitude • not needed to describe choice behavior 29 / 43 Utility Two concepts of utility: Cardinal utility – attach a significance to the magnitude of utility: • difficult to assign the magnitude • not needed to describe choice behavior Ordinal utility – important is only the order of preference: • easy to set the utility – 1 rule: preferred bundle has a higher utility • we can derive a complete theory of demand We will use the ordinal utility. 29 / 43 Ordinal utility Utility function is a way of assigning a number to every possible consumption bundle such that more-preferred bundles get assigned larger numbers than less-preferred bundles. If (x1, x2) (y1, y2), then u(x1, x2) > u(y1, y2). 30 / 43 Ordinal utility Utility function is a way of assigning a number to every possible consumption bundle such that more-preferred bundles get assigned larger numbers than less-preferred bundles. If (x1, x2) (y1, y2), then u(x1, x2) > u(y1, y2). Different ways to assign utilities that describe the same preferences: 30 / 43 Monotonic transformation Positive monotonic transformation f (u) = any increasing function of u. Describes the same preferences as the original utility function u. Examples of the function f (u): f (u) = 3u, f (u) = u + 3, f (u) = u3 31 / 43 Monotonic transformation Positive monotonic transformation f (u) = any increasing function of u. Describes the same preferences as the original utility function u. Examples of the function f (u): f (u) = 3u, f (u) = u + 3, f (u) = u3 Example: Two bundles X and Y , preferences: X Y We assign utility so that u(X) > u(Y ), e.g. u(X) = 1, u(Y ) = −1 Do monotonic transformations f1(u) = 3u a f2(u) = u + 3 represent the same preferences as the original utility function u? 31 / 43 Monotonic transformation Positive monotonic transformation f (u) = any increasing function of u. Describes the same preferences as the original utility function u. Examples of the function f (u): f (u) = 3u, f (u) = u + 3, f (u) = u3 Example: Two bundles X and Y , preferences: X Y We assign utility so that u(X) > u(Y ), e.g. u(X) = 1, u(Y ) = −1 Do monotonic transformations f1(u) = 3u a f2(u) = u + 3 represent the same preferences as the original utility function u? Yes: • f1(u) = 3u: f1(u(X)) = 3 > −3 = f1(u(Y )) • f2(u) = u + 3: f2(u(X)) = 4 > 2 = f2(u(Y )) 31 / 43 Construction of indifference curves from utility function Utility function u(x1, x2) = x1x2 32 / 43 Construction of indifference curves from utility function Utility function u(x1, x2) = x1x2 =⇒ indifference curves x2 = k x1 32 / 43 PROBLEM: The slope of indifference curves The slope of indifference curves for two utility functions: 1. What is the slope of IC x2 = 4/x1 v point (x1, x2) = (2, 2)? 33 / 43 PROBLEM: The slope of indifference curves The slope of indifference curves for two utility functions: 1. What is the slope of IC x2 = 4/x1 v point (x1, x2) = (2, 2)? Slope of indifference curves = MRS = dx2 dx1 = −4 x2 1 = −1 33 / 43 PROBLEM: The slope of indifference curves The slope of indifference curves for two utility functions: 1. What is the slope of IC x2 = 4/x1 v point (x1, x2) = (2, 2)? Slope of indifference curves = MRS = dx2 dx1 = −4 x2 1 = −1 2. What is the slope of IC x2 = 10 − 6 √ x1 v point (4, 5)? 33 / 43 PROBLEM: The slope of indifference curves The slope of indifference curves for two utility functions: 1. What is the slope of IC x2 = 4/x1 v point (x1, x2) = (2, 2)? Slope of indifference curves = MRS = dx2 dx1 = −4 x2 1 = −1 2. What is the slope of IC x2 = 10 − 6 √ x1 v point (4, 5)? Slope of indifference curves = MRS = dx2 dx1 = −3 √ x1 = −3 2 33 / 43 Examples of utility functions – perfect substitutes The consumer is willing to exchange • coke and pepsi at a ratio 1:1 34 / 43 Examples of utility functions – perfect substitutes The consumer is willing to exchange • coke and pepsi at a ratio 1:1 important is the total number: e.g. u(K, P) = K + P 34 / 43 Examples of utility functions – perfect substitutes The consumer is willing to exchange • coke and pepsi at a ratio 1:1 important is the total number: e.g. u(K, P) = K + P • 2 buns for 1 baguette 34 / 43 Examples of utility functions – perfect substitutes The consumer is willing to exchange • coke and pepsi at a ratio 1:1 important is the total number: e.g. u(K, P) = K + P • 2 buns for 1 baguette baguette has a double weight: e.g. u(R, H) = R + 2H 34 / 43 Examples of utility functions – perfect complements The consumer demands • left and right shoes at a fixed ratio 1:1 35 / 43 Examples of utility functions – perfect complements The consumer demands • left and right shoes at a fixed ratio 1:1 lower quantity matters: e.g. u(L, P) = min{L, P} 35 / 43 Examples of utility functions – perfect complements The consumer demands • left and right shoes at a fixed ratio 1:1 lower quantity matters: e.g. u(L, P) = min{L, P} • rum and coke at a fixed ratio 1:5 35 / 43 Examples of utility functions – perfect complements The consumer demands • left and right shoes at a fixed ratio 1:1 lower quantity matters: e.g. u(L, P) = min{L, P} • rum and coke at a fixed ratio 1:5 goal: same numbers in the bracket – we need only 1/5 of coke: e.g. u(R, K) = min{5R, K} 35 / 43 Examples of utility functions – quasilinear preferences Indifference curves are vertically parallel (a practical property) Utility function u(x1, x2) = v(x1) + x2, e.g. u(x1, x2) = √ x1 + x2 36 / 43 Examples of utility functions – Cobb-Douglas preferences • A simple utility function representing well-behaved preferences. • Utility function of the form u(x1, x2) = xc 1 xd 2 . • More convenient to use the transformation f (u) = u 1 c+d and write xa 1 x1−a 2 , where a = c/(c + d). 37 / 43 Examples of utility functions – Cobb-Douglas preferences • A simple utility function representing well-behaved preferences. • Utility function of the form u(x1, x2) = x1 1 x2 2 . • More convenient to use the transformation f (u) = u1/3 and write x 1/3 1 x 2/3 2 . 37 / 43 Marginal utility Marginal utility (MU) is the change in utility from an increase in consumption of one good, while the quantities of other goods are constant. Partial derivatives of u(x1, x2) with respect to x1 or x2. Pˇr´ıklady: • u(x1, x2) = x1 + x2 → MU1 = ∂u/∂x1 = 1 • u(x1, x2) = xa 1 x1−a 2 → MU2 = ∂u/∂x2 = (1 − a)xa 1 x−a 2 38 / 43 Marginal utility Marginal utility (MU) is the change in utility from an increase in consumption of one good, while the quantities of other goods are constant. Partial derivatives of u(x1, x2) with respect to x1 or x2. Pˇr´ıklady: • u(x1, x2) = x1 + x2 → MU1 = ∂u/∂x1 = 1 • u(x1, x2) = xa 1 x1−a 2 → MU2 = ∂u/∂x2 = (1 − a)xa 1 x−a 2 The value of MU changes with a monotonic transformation of the utility function. If we multiply utility times 2, MU increases times 2. 38 / 43 Relationship between MU and MRS We want to measure MRS = slope of IC u(x1, x2) = k, where k is a constant. We are interested in (∆x1, ∆x2), for which the utility is constant: MU1∆x1 + MU2∆x2 = 0 MRS = ∆x2 ∆x1 = − MU1 MU2 We can calculate MRS from the utility function. E.g. for u = √ x1x2: MRS = − ∂u/∂x1 ∂u/∂x2 = − 0,5x−0,5 1 x0,5 2 0,5x0,5 1 x−0,5 2 = − x2 x1 39 / 43 Relationship between MU and MRS We want to measure MRS = slope of IC u(x1, x2) = k, where k is a constant. We are interested in (∆x1, ∆x2), for which the utility is constant: MU1∆x1 + MU2∆x2 = 0 MRS = ∆x2 ∆x1 = − MU1 MU2 We can calculate MRS from the utility function. E.g. for u = √ x1x2: MRS = − ∂u/∂x1 ∂u/∂x2 = − 0,5x−0,5 1 x0,5 2 0,5x0,5 1 x−0,5 2 = − x2 x1 The value of MRS does not change with monotonic transformation. If we multiply utility function times 2, MRS= −2MU1 2MU2 = −MU1 MU2 . 39 / 43 APPLICATION: Utility from commuting People decide whether to take bus or car. Each type of transport represents a bundle with different characteristics, e.g.: • x1 is walking time • x2 is time taking a bus or car • x3 is the total cost of commuting • ... Assume that the utility function has a linear form U(x1, ..., xn) = β1x1 + ... + βnxn. Then we use statistical techniques to estimate the parameters βi that best describe choices. 40 / 43 APPLICATION: Utility from commuting (cont’d) Domenich and McFadden (1975) estimated the following utility function: U(TW , TT, C) = −0,147TW − 0,0411TT − 2,24C • TW = total walking time in minutes • TT = total driving time in minutes • C = total cost in dollars 41 / 43 APPLICATION: Utility from commuting (cont’d) Domenich and McFadden (1975) estimated the following utility function: U(TW , TT, C) = −0,147TW − 0,0411TT − 2,24C • TW = total walking time in minutes • TT = total driving time in minutes • C = total cost in dollars The parameters can be used for different purposes. For instance, we can: • calculate the marginal rate of substitution between two characteristics • forecast consumer response to proposed changes • estimate whether a change is worthwhile in a benefit-cost sense 41 / 43 What should you know? • Budget set = consumption bundles available at given prices and income • Budget line are bundles for which the entire income is spent. • If the preference relation is complete, reflexive and transitive, consumer can order bundles according to preferences. • Monotonicity and convexity are reasonable assumptions – easier to find the optimum bundle. 42 / 43 What should you know? (cont’d) • Utility function assigns numbers to different bundles so that the bundles are ordered according to preferences. • The numbers have no meaning in itself. Monotonic transformation of u represents the same preferences.. • MRS measures the slope of IC. • The slope of IC measures the willingness to pay for good 1 (in units of good 2) • The slope of BL measures the opportunity cost of good 1(in units of good 2) 43 / 43