Thursday, June 16, 2016

Cyclical preferences and the exchange economy

Some experimentation on assumptions of competitive equilibrium yields very interesting results. Suppose in an exchange economy, the first individual has a year long utility function:

(1) U = X1*Y1

And the second individual:

(2) Z1= X2*Y2 during summertime

(3)  Z2=X2*(Y2)^2 during wintertime

Suppose the first individual was endowed with the basket w1(90,80), and the second individual  w2(10,20).

Solving for X1,Y1,X2,Y2 where individuals maximize their utility given their constraints each season gives the following table:

Years Season X1 Y1 X2 Y2 Py/Px Px/Py
Endowement 90 80 10 20

1 Summer 85.00 85.00 15.00 15.00 1.0000 1.0000
Winter 80.95 89.47 19.05 10.53 0.9048 1.1053
2 Summer 85.21 85.21 14.79 14.79 1.0000 1.0000
Winter 81.21 89.63 18.79 10.37 0.9061 1.1037
3 Summer 85.42 85.42 14.58 14.58 1.0000 1.0000
Winter 81.46 89.78 18.54 10.22 0.9073 1.1022
4 Summer 85.62 85.62 14.38 14.38 1.0000 1.0000
Winter 78.82 93.71 21.18 6.29 0.8412 1.1888
5 Summer 86.26 86.26 13.74 13.74 1.0000 1.0000
Winter 79.70 94.01 20.30 5.99 0.8477 1.1796
6 Summer 86.85 86.85 13.15 13.15 1.0000 1.0000
Winter 80.50 94.29 19.50 5.71 0.8538 1.1713

(Py/Px is a price ratio)

While it is not possible to compare allocations during Summer and Winter seasons, we can compare Summers of the each year. The second individual who has cyclical preferences is worse off, not only he was unlucky to be poor at  the beginning, but also he is getting poorer in consequent years. Inequality between them gradually expanding. Nevertheless, all points are Pareto efficient.

Is it realistic to assume that the second individual has different preferences during different seasons? One explanation could be that people's needs during summer are different than in winter, and they frequently trade their goods to match their urgent needs in both periods. Especially if they are poor.

It is also curious that once preference of the individual has shifted, it is impossible to get to the original basket anymore. Given that utility function is Cobb-Douglas type and that other individual's preference does not change.


        (each dot represents X1, Y1 allocation in particular season,  they are approaching to 100 ).