Intermediate Micro

• Workhorse model of intermediate micro

– Utility maximization problem
Expected Utility – Consumers Max U(x,y) subject to the budget constraint,
I=Pxx + Pyy

Health Economics • Problem is made easier by the fact that we assume all
Fall 2018 variables are known with certainty
– Consumers know prices and income
– Know exactly the quality of the product

1 2

• Many cases, there is uncertainty about some variables • Will emphasize the special role of insurance in a generic
– Uncertainty about income? sense
– What are prices now? What will prices be in the future? – Why insurance is ‘good’ ?
– How much insurance should people purchase?
– Uncertainty about quality of the product?
– Compare that to how insurance is usually structured in health
care
• This section, will review utility theory under uncertainty

3 4

1
 Special problems of health care insurance Definitions

• Moral hazard • Probability - likelihood discrete event will occur

– Reimbursement structure of health insurance encourages – n possible events, i=1,2,..n
more use of medical care – Pi be the probability event i happens
– 0 ≤Pi≤1
• Adverse selection – P1+P2+P3+…Pn=1
– Those with the most needs for medical care are attracted to • Probabilities can be ‘subjective’ or ‘objective’,
insurance depending on the model
• In our work, probabilities will be know with certainty
• What these problems do to markets?

• What these problems due to welfare? 5 6

• Expected value – • Roll of a die, all sides have (1/6) prob. What is
– Weighted average of possibilities, weight is probability expected roll?
– Sum of the possibilities times probabilities
• E(x) = 1(1/6) + 2(1/6) + … 6(1/6) = 3.5
• x={x1,x2…xn}
• P={P1,P2,…Pn} • Suppose you have: 25% chance of an A, 50% B, 20%
C, 4% D and 1% F
• E(x) = P1X1 + P2X2 + P3X3 +….PnXn • E[quality points] = 4(.25) + 3(.5) + 2(.2) + 1(.04) +
0(.01) = 2.94
7 8

2
 Expected utility

• Suppose income is random. Two potential values (Y1 • However, suppose an agent is faced with choice
or Y2) between two different paths
– Choice a: Y1 with probability P1 and Y2 with P2
• Probabilities are either P1 or P2=1-P1 – Choice b: Y3 with probability P3 and Y4 with P4

• When incomes are realized, consumer will experience a • Example: You are presented with two option
particular level of income and hence utility – a job with steady pay or
– a job with huge upside income potential, but one with a
chance you will be looking for another job soon
• But, looking at the problem beforehand, a person has a
particular ‘expected utility’
• How do you choose between these two options?
9 10

Utility
Assumptions about utility with uncertainty
U = f(Y)

• Utility is a function of one element (income or wealth),

where U = U(Y) U1

• Marginal utility is positive

– U' = dU/dY > 0

• Standard assumption, declining marginal utility U ' ' <0

– Implies risk averse but we will relax this later

Y1 Income
11 12

3
Utility Utility
Ub U = f(Y) U = f(Y)
U2 U2

Ua
U1 U1

Y1 Y1+a Y2 Y2+a Income Y1 Y2 Income

13 14

Von Neumann-Morganstern Utility

• N states of the world, with incomes defined as Y1 Y2 • E(U) is the sum of the possibilities times probabilities
….Yn
• Example:
• The probabilities for each of these states is P1 P2…Pn – 40% chance of earning $2500/month
– 60% change of $1600/month
– U(Y) = Y0.5
• A valid utility function is the expected utility of the
– Expected utility
gamble
• E(U) = P1U(Y1) + P2U(Y2)
• E(U) = 0.4(2500)0.5 + 0.6(1600)0.5
• E(U) = P1U(Y1) + P2U(Y2) …. + PnU(Yn) = 0.4(50) + 0.6(40) = 44

15 16

4
 Example

• Note that expected utility in this case is very different • Job A: certain income of $50K
from expected income
– E(Y) = 0.4(2500) + 0.6(1600) = 1960 • Job B: 50% chance of $10K and 50% chance of $90K

• Expected utility allows people to compare gambles • Expected income is the same ($50K) but in one case,
income is much more certain
• Given two gambles, we assume people prefer the
situation that generates the greatest expected utility • Which one is preferred?
– People maximize expected utility

17 18

Another Example

• U=ln(y) • Job 1
– 40% chance of $2500, 60% of $1600
• EUa = ln(50,000) = 10.82 – E(Y1) = 0.4*2500 + .6*1600 = $1960
– E(U1) = (0.4)(2500)0.5 + (0.6)(1600)0.5 =44
• EUb = 0.5 ln(10,000) + 0.5ln(90,000) = 10.31 • Job 2
– 25% chance of $5000, 75% of $1000
• Job (a) is preferred – E(Y2) = .25(5000) + .75(1000) = $2000
– E(U2) = 0.25(5000)0.5 + 0.75(1000)0.5 = 41.4
• Job 1 is preferred to 2, even though 2 has higher
19 expected income 20

5
 The Importance of Marginal Utility:
The St Petersburg Paradox
• Bet starts at $2. Flip a coin and if a head appears, the • Probabilities?
bet doubles. If tails appears, you win the pot and the
game ends. • Pr(h)=Pr(t) = 0.5

• All events are independent

• So, if you get H, H, H T, you win $16
• Pr(h on 2nd | h on 1st) = Pr(h on 2nd)

• What would you be willing to pay to ‘play’ this game? • Recall definition of independence
• If A and B and independent events
– Pr (A ∩ B) = Pr(A)Pr(B)

21 22

• Note, Pr(first tail on kth toss) = • What is the expected value of the gamble
• Pr(h on 1st)Pr(h on 2nd …)…Pr(t on kth)=
• (1/2)(1/2)….(1/2) = (1/2)k • E = (1/2)$21 + (1/2)2$22 + (1/2)3$23 + (1/2)4$24

• What is the expected pot on the kth trial? k

 
1
• 2 on 1st or 21
 
E     2k   1  
• 4 on 2nd or 22 k 1  2  k 1
• 8 on 3rd, or 23
• So the payoff on the kth is 2k • The expected payout is infinite

23 24

6
 Suppose Utility is U=Y0.5? What is E[U]?
 k  k /2 k /2
1 1 1
E     2k      2 
1/2 1/2
Round Winnings Probability k
 
k 1  2  k 1  2  2
5th $32 0.03125

10th $1,024 0.000977 

1
k /2
1
k /2 
1
k /2 
 1 
k

    2     
k /2
   
k 1  2  2 k 1  2  k 1  2
15th $32,768 3.05E-5

20th $1,048,576 9.54E-7  

 1  1 
k


25th $33,554,432 2.98E-8 Can show that         1  2.414

k 1  2   1 1 
 
 2
25 26

How to represent graphically Utility

U1 U(Y)
b
• Probability P1 of having Y1
• (1-P1) of having Y2 U3

• U1 and U2 are utility that one would receive if they

U4
received Y1 and Y2 respectively c
• E(Y) =P1Y1 + (1-P1)Y2 = Y3
• U3 is utility they would receive if they had income Y3 U2 a
with certainty

Y2 Y3=E(Y) Income
Y1
27 28

7
• Notice that E(U) is a weighted average of utilities in the • Draw vertical line from E(Y) to the line segment. This
good and bad states of the world is E(U)
• E(U) = P1U(Y1) + (1-P1)U(Y2) • U4 is Expected utility
• The weights sum to 1 (the probabilities) • U4 = E(U) = P1U(Y1) + (1-P1)U(Y2)
• Draw a line from points (a,b)
• Represent all the possible ‘weighted averages’ of U(Y1)
and U(Y2)
• What is the one represented by this gamble?

29 30

Numeric Example

• Suppose offered two jobs • Job A

– Job A: Has chance of a high (Y1) and low (Y2) wages – 20% chance of $150,000
– Job B: Has chance of high (Y3) and low (Y4) wages – 80% chance of $20,000
– Expected income from both jobs is the same – E(Y) = 0.2(150K) + 0.8(20K) = $46K
– Pa and Pb are the probabilities of getting the high wage
situation • Job B
– 60% chance of $50K
PaY1 + (1-Pa)Y2 = PbY3 + (1-Pb)Y4 =E(Y) – 40% chance of $40K
– E(Y) = 0.6(50K) + 0.4(40K) = $46K

31 32

8
Utility
U(Y)

• Notice that Job A and B have the same expected

income
Ub
• Job A is riskier – bigger downside for Job A
Ua
• Prefer Job B (Why? Will answer in a moment)

Y2 E(Y) Income
Y4 Y3 Y1
33 34

Example

• The prior example about the two jobs is instructive. • Suppose have $200,000 home (wealth).
Two jobs, same expected income, very different • Small chance that a fire will damage you house. If does,
expected utility will generate $75,000 in loss (L)
• People prefer the job with the lower risk, even though • U(W) = ln(W)
they have the same expected income • Prob of a loss is 0.02 or 2%
• People prefer to ‘shed’ risk – to get rid of it. • Wealth in “good” state = W
• How much are they willing to pay to shed risk? • Wealth in bad state = W-L

35 36

9
• E(W) = (1-P)W + P(W-L) • Suppose you can add a fire detection/prevention
• E(W) = 0.98(200,000) + 0.02(125,000) = $198,500 system to your house.
• This would reduce the chance of a bad event to 0 but it
• E(U) = (1-P) ln(W) + P ln(W-L) would cost you $C to install
• E(U) = 0.98 ln(200K) + 0.02 ln(200K-75K) = 12.197 • What is the most you are willing to pay for the security
system?
• E(U) in the current situation is 12.197
• Utility with the security system is U(W-C)
• Set U(W-C) equal to 12.197 and solve for C
37 38

Utility

U1 U(W)
b
• ln(W-C) =12.197
• Recall that eln(x) = x Risk Premium

• Raise both sides to the e d

U4
• eln(W-C) = W-C = e12.197 = 198,128 c

• 198,500 – 198,128 = $372

U2 a
• Expected loss is $1500
• Would be willing to pay $372 to avoid that loss
W-L Y3=E(W) Wealth
Y4 W
39 40

10
• Will earn Y1 with probability p1 • Take the expected income, E(Y). Draw a line to (ab).
– Generates utility U1 The height of this line is E(U).
• Will earn Y2 with probability p2=1-p1 • E(U) at E(Y) is U4
– Generates utility U2 • Suppose income is know with certainty at I3. Notice
• E(I) =p1Y1 + (1-p1)Y2 = Y3 that utility would be U3, which is greater that U4
• Line (ab) is a weighted average of U1 and U2 • Look at Y4. Note that the Y4<Y3=E(Y) but these two
• Note that expected utility is also a weighted average situations generate the same utility – one is expected,
one is known with certainty
• A line from E(Y) to the line (ab) give E(U) for given
E(Y)

41 42

Some numbers

• The line segment (cd) is the “Risk Premium.” It is the • Person has a job that has uncertain income
amount a person is willing to pay to avoid the risky – 50% chance of making $30K, U(30K) = 18
situation. – 50% chance of making $10K, U(10K) = 10
• If you offered a person the gamble of Y3 or income Y4, • Another job with certain income of $16K
they would be indifferent. – Assume U($16K)=14
• Therefore, people are willing to sacrifice cash to ‘shed’
risk. • E(I) = (0.5)($30K) + (0.05)($10K) = $20K
• E(U) = 0.5U(30K) + 0.5U(10K) = 14

43 44

11
 Utility
18 U = f(I)
b
• Expected utility. Weighted average of U(30) and U(10). 16
E(U) = 14
14 c
• Notice that a gamble that gives expected income of d
$20K is equal in value to a certain income of only $16K
• This person dislikes risk. 10 a
– Indifferent between certain income of $16 and uncertain
income with expected value of $20
– Utility of certain $20 is a lot higher than utility of uncertain
income with expected value of $20

$10 $16 $20 $30 Income

45 46

• Although both jobs provide the same expected income, • Notice also that the person is indifferent between a job
the person would prefer the guaranteed $20K. with $16K in certain income and $20,000 in uncertain
• Why? Because of our assumption about diminishing • They are willing to sacrifice up to $4000 in income to
marginal utility reduce risk, risk premium
– In the ‘good’ state of the world, the gain from $20K to $30K
is not as valued as the 1st $10
– In the ‘bad’ state, because the first $10K is valued more than
the last $10K, you lose lots of utils.

47 48

12
 Example Utility
30 U = f(I)
b
• U= y0.5
• Job with certain income c
18
– $400 week d
– U=4000.5=20
$76
• Can take another job that 10 a
– 40% chance of $900/week, U=30
– 60% chance of $100/week, U=10
– E(I) = 420, E(U) = 0.4(30) + 0.6(10) = 18

$100 $324 $420 $900 Income

49 50

Risk Loving

• Notice that utility from certain income stream is higher • The desire to shed risk is due to the assumption of
even though expected income is lower declining marginal utility of income
• What is the risk premium?? • Consider the next situation.
• What certain income would leave the person with a • The graph shows increasing marginal utility of income
utility of 18? U=Y0.5
• So if 18 = Y0.5, 182= Y =324 • U`(Y1) > U`(Y2) even though Y1>Y2
• Person is willing to pay 400-324 = $76
to avoid moving to the risky job

51 52

13
Utility Utility
U = f(Y) U = f(Y)

U1

U3

U4

U2
Income Income
Y1 Y2 53
Y2 Y3=E(Y) Y1 54

Risk Neutral

• What does this imply about tolerance for risk? • If utility function is linear, the marginal utility of
• Notice that at E(Y) = Y3, expected utility is U3. income is the same for all values of income
– U ' >0
• Utility from a certain stream of income at Y3 would – U ' ' =0
generate U4. Note that U3>U4 • The uncertain income E(Y) and the certain income Y3
• This person prefers an uncertain stream of Y3 instead generate the same utility
of a certain stream of Y3 • This person is considered risk neutral
• This person is ‘risk loving’. Again, the result is driven • We usually make the assumption firms are risk neutral
by the assumption are U``

55 56

14
 Example Utility
U = a+bY

• 25% chance of $100

• 75% chance of $1000 U1
• E[Y] = 0.25(100) + 0.75(1000) = $775

• U=Y
U3 =U4

• Compare to certain stream of $775

U2

Income
57
Y2 Y3=E(y) Y1 58

Benefits of insurance Simple insurance example

• Assume declining marginal utility • Suppose income is know (Y1) but random --shocks can
• Person dislikes risk reduce income
– House or car is damaged
– They are willing to receive lower certain income rather than
– Can pay $ to repair, return you to the normal state of world
higher expected income
• L is the loss if the bad event happens
• Firms can capitalize on the dislike for risk by helping
people shed risk via insurance • Probability of loss is P1
• Expected utility without insurance is
• E(U) = (1-P1)U(Y1) + P1U(Y1-L)

59 60

15
• Suppose you can buy insurance that costs you PREM. • Notice that insurance has made income certain. You
The insurance pay you to compensate for the loss L. will always have income of Y-PREM
– In good state, income is • What is the most this person will pay for insurance?
• Y-Prem • The expected loss is p1L
– In bad state, paid PREM, lose L but receive PAYMENT,
therefore, income is • Expected income is E(Y)
• Y-Prem-L+Payment • The expected utility is U2
– For now, lets assume PAYMENT=L, so • People would always be willing to pay a premium that
– Income in the bad state is also equaled the expected loss
• Y-Prem

61 62

Utility

• But they are also willing to pay a premium to shed risk

(line cd) U2 d
c
• The maximum amount they are willing to pay is
expected loss + risk premium
Willingness
to pay for
insurance

Y-L Y2 E(Y) Income

Y
63 64

16
• Suppose income is $50K, and there is a 5% chance of • E(U) = P ln(Y-L) + (1-P)ln(Y)
having a car accident that will generate $15,000 in loss
• Expected loss is .05(15K) = $750
• E(U) = 0.05 ln(35,000) + 0.95 ln(50,000)
• U = ln(y)
• Some properties of logs
Y=ln(x) then ey = exp(y) = x • E(U) = 10.8
Y=ln(xa) = a ln(x)
Y=ln(xz) = ln(x) + ln(z)

65 66

• What is the most someone will pay for insurance? • Recall that the expected loss is $750 but this person is
• People would purchase insurance so long as utility with willing to pay more than the expected loss to avoid the
certainty is at least 10.8 (expected utility without risk
insurance) • Pay $750 (expected loss), plus the risk premium ($979-
• Ua =U(Y – Prem) ≥ 10.8 $750) = 229
• Ln(Y-PREM) ≥10.8
• Y-PREM = exp(10.8)
• PREM =Y-exp(10.8) = 50,000 – 49,021 =979

67 68

17
Utility
Supply of Insurance

• Suppose there are a lot of people with the same

situation as in the previous slide
U2 d
c • Each of these people have a probability of loss P and
when a loss occurs, they have L expenses
$229
• A firm could collect money from as many people as
possible in advance. If bad event happens, they pay
back a specified amount.

$35,000 $49,021 $49,250 $50,000 Income

69 70

• Firms are risk neutral, so they are interested in expected • Think of the profits made on sales to one person
profits
• Expected profits = revenues – costs • A person buys a policy that will pay them q dollars
– Revenues are known (q≤L) back if the event occurs
– Some of the costs are random (e.g., exactly how many claims
you will pay) • To buy this insurance, person will pay “a” dollars per
dollar of coverage

• Cost per policy is fixed t

71 72

18
• Revenues = aq • a = p + (t/q)
– a is the price per dollar of coverage
• The cost per dollar of coverage is proportion to risk
• Costs =pq +t
– For every dollar of coverage (q) expect to pay this p percent • t/q is the loading factor. Portion of price to cover
of time administrative costs
• E(π) = aq – pq – t • Make it simple, suppose t=0.
• Let assume a perfectly competitive market, so in the long run π
=0 – a=p
• What should the firm charge per dollar of coverage? – If the probability of loss is 0.05, will change 5 cents per $1.00
• E(π) = aq – pq – t = 0 of coverage

73 74

How much insurance will people purchase when prices

are actuarially fair?

• In this situation, if a person buys a policy to insure L • With insurance

dollars, the ‘actuarially fair’ premium will be LP – Pay a premium that is subtracted from income
• An actuarially fair premium is one where the premium – If bad state happens, lose L but get back the amount of
equals the expected loss insurance q
– They pay p+(t/q) per dollar of coverage. Have q dollars of
• In the real world, no premiums are ‘actuarially fair’
coverage – so they to pay a premium of pq+t in total
because prices include administrative costs called
‘loading factors’ • Utility in good state
– U = U[Y – pq - t]

75 76

19
• Utility in bad state • E(u) = (1-p)U[Y – pq] + pU[Y-L+q-pq]
– U[Y- L + q – pq - t]
• dE(u)/dq = (1-p) U'(y-pq)(-p)
• E(u) = (1-p)U[Y – pq – t] + pU[Y-L+q-pq-t] • + pU'(Y-L+q-pq)(1-p) = 0
• Simplify, let t=0 (no loading costs)
• E(u) = (1-p)U[Y – pq] + pU[Y-L+q-pq] • p(1-p)U'(Y-L+q-pq) = (1-p)pU'(Y-pq)
• Maximize utility by picking optimal q
• (1-p)p cancel on each side
• dE(u)/dq = 0

77 78

Insurance w/ loading costs

• U'(Y-L+q-pq) = U'(Y-pq) • Insurance is not actuarially fair and insurance does have
• Optimal insurance is one that sets marginal utilities in loading costs
the bad and good states equal • Can show (but more difficult) that with loading costs,
• Y-L+q-pq = Y-pq people will now under-insure, that is, will insure for less
• Y’s cancel, pq’s cancel, than the loss L
• q=L • Intution? For every dollar of expected loss you cover,
• If people can buy insurance that is ‘fair’ they will fully will cost more than a $1
insure loses. • Only get back $1 in coverage if the bad state of the
world happens

79 80

20
• Recall: • E(u) = (1-p)U[Y – pqk] + pU[Y-L+q-pqk]
– q is the amount of insurance purchased
– Without loading costs, cost per dollar of coverage is p • dE(u)/dq = (1-p) U' (y-pqk)(-pk)
– Now, for simplicity, assume that price per dollar of coverage
is pK where K>1 (loading costs)
• + pU'(Y-L+q-pqk)(1-pk) = 0
• Buy q $ worth of coverage
• Pay qpK in premiums • p(1-pk)U'(Y-L+q-pqk) = (1-p)pkU'(Y-pqk)

• p cancel on each side

81 82

• (1-pk)U'(Y-L+q-pkq) = (1-p)kU' (Y-pkq) • (Y-L+q-pqk) < (Y-pqk)

• (a)(b) = (c)(d) • Y and –pqk cancel
• Since k > 1, can show that • -L + q < 0
• (1-pk) < (1-p)k • Which means that q < L
• Since (a) < (c), must be the case that • When price is not ‘fair’ you will not fully insure
• (b) > (d)
• U'(Y-L+q-pkq) > U'(Y-pkq)
• Since U'(y1) > U'(y2), must be that y1 < y2

83 84

21
 Demand for Insurance

• Both people have income of Y • Probabilities the health shock will occur are P1 and P2
• Each person has a potential health shock • Expected Income of person 1
– The shock will leave person 1 w/ expenses of E1 and will – E(Y)1 = (1-P1)Y + P1*(Y-E1)
leave income at Y1=Y-E1 – E(Y)2 = (1-P2)Y + P2*(Y-E2)
– The shock will leave person 2 w/ expenses of E2 and will – Suppose that E(Y)1 = E(Y)2 = Y3
leave income at Y2=Y-E2
• Suppose that
– E1>E2, Y1<Y2

85 86

Utility
a U(Y)
Ub g
• In this case c f
– Shock 1 is a low probability/high cost shock Ua
d
– Shock 2 is a high probability/low cost shock
• Example
– Y=$60,000
– Shock 1 is 1% probability of $50,000 expense
– Shock 2 is a 50% chance of $1000 expense
– E(Y) = $59500 b

Y1 Ya Y2 E(Y)=Y3 Y Income
88
87
Yb

22
 Implications

• Expected utility locus • Do not insure small risks/high probability events

– Line ab for person 1 – If you know with certainty that a costs will happen, or, costs
– Line ac for person 2 are low when a bad event occurs, then do not insure
• Expected utility is – Example: teeth cleanings. You know they happen twice a
– Ua in case 1 year, why pay the loading cost on an event that will happen?
– Ub in case 2
• Certainty premium –
– Line (de) for person 1, Difference Y3 – Ya
– Line (fg) for person 2, Difference Y3 - Yb

89 90

Some adjustments to this model

• Insure catastrophic events • The model assumes that poor health has a monetary
– Large but rare risks cost and that is all.
– When experience a bad health shock, it costs you L to
• As we will see, many of the insurance contracts we see recover and you are returned to new
do not fit these characteristics – they pay for small • Many situations where
predictable expenses and leave exposed catastrophic – health shocks generate large expenses
events – And the expenses may not return you to normal
– AIDS, stroke, diabetes, etc.

91 92

23
• In these cases, the health shock has fundamentally • Typical assumption
changed life. – U(Y) >V(Y)
• We can deal with this situation in the expected utility • For any given income level, get higher utility in the
model with adjustment in the utility function healthy state
– U`(Y) > V`(Y)
• “State dependent” utility
• For any given income level, marginal utility of the next
– U(y) utility in healthy state
dollar is higher in the healthy state
– V(y) utility in unhealthy state

93 94

Utility Note that:

a
U(y)

a • At Y1,
c b – U(Y1) > V(Y1)
– U`(Y1) > V`(Y1)
b V(y)
– Slope of line aa > slope of line bb
c
• Notice that slope line aa = slope of line cc
– U`(Y1) = V`(Y2)

Income
Y2 Y1 Y
95 96

24
 What does this do to optimal insurance

• E(u) = (1-p)U[Y – pq – t] + pV[Y-L+q-pq-t] • Just like in previous case, we equalize marginal utility
• Again, lets set t=0 to make things easy across the good and bad states of the world
• Recall that
• E(u) = (1-p)U[Y – pq] + pV[Y-L+q-pq]
– U`(y) > V`(y)
– U`(y1) = V`(y2) if y1>y2
• dE(u)/dq = (1-p)(-p)U`[Y-pq] • Since U`[Y-pq] = V`[Y-l+q-pq]
+p(1-p)V`[Y-l+q+pq] = 0 • In order to equalize marginal utilities of income, must
• U`[Y-pq] = V`[Y-l+q-pq] be the case that
[Y-pq] > [Y-l+q+pq]

97 98

Allais Paradox

• Income in healthy state > income in unhealthy state • Which gamble would you prefer
• Do not fully insure losses. Why? – 1A: $1 million w/ certainty
– With insurance, you take $ from the good state of the world – 1B: (.89, $1 million), (0.01, $0), (0.1, $5 million)
(where MU of income is high) and transfer $ to the bad state
of the world (where MU is low) • Which gamble would you prefer
– Do not want good money to chance bad – 2A: (0.89, $0), (0.11, $1 million)
– 2B: (0.9, $0), (0.10, $5 million)

99 100

25
• 1st gamble:
• U(1) > 0.89U(1) + 0.01U(0) + 0.1U(5)
• 0.11U(1) > 0.01U(0) + 0.1U(5)

• Now consider gamble 2

• 0.9U(0) + 0.1U(5) > 0.89U(0) + 0.11U(1)
• 0.01U(0) + 0.1U(5) > 0.11U(1)

• Choice of Lottery 1A and 2B is inconsistent with

expected utility theory

101

26