The Experimental Determination of Indifference Curves

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 3

The Experimental Determination of Indifference

curves
Introduction

The Author discusses about the empirical derivation of Indifference curves and provides insights into
the work done so far and also proposes a method to carry out a procedure to experimentally
determine indifference curves. The author emphasise that empirical determination of indifference
curve is useful for determining preferences in non-market conditions apart from determining consumer
demand.
The author discusses about the merits of determination of indifference curve based upon actual
expenditure and experimental data. Though indifference curves obtained from actual data would be
more convincing, however the constraints such as number of economic units observed ,time period of
observation, varying situations affect the outcome.
Through this paper , the author based on the learnings of two previous studies by Thurstone and
another by Rousseas and Hart, proposes a new procedure to experimentally determine the
indifference curves.

Process to experimentally determine Indifference curve

Previous Studies - The two previous studies by Thurstone and another by Rousseas and Hart
conducted and were reviewed by the author .The key points noted by the author about these
experiments are as follows-

1. Thurstone’s Experiment
a. Thurstone was the first to experimentally determine indifference curves.
b. Provided a useful feature about comparing commodities bundles with reference
bundles.
c. Shortcomings as observed by Author -
i. Commodities selected for experimentation were not typically required in large
nos and hence the subjects choice was purely hypothetical.
ii. No of comparisons required by the subject to make the choices were very
large.
iii. Assumption about the shape of indifference curve.
iv. There was only one subject.

2. Rousseas and Hart Experiment


a. No of subjects utilised was fairly large – 67
b. Experiment was conducted twice within a month difference thus generating
preference data at two distinct time periods.
c. Provided a useful feature of forced consumption of the commodities that contributed
to a more careful consideration by the subjects thus ensuring a closer to real life
situation.
d. Shortcomings as observed by the Author –
i. Assumed homogeneity of tastes for the commodities selected i.e eggs and
bacon.
ii. Procedure adopted to fit curves was found to be cumbersome.

Criticism about the usefulness of Indifference curve analysis on empirical data by Wallis and
Friedman

Basis the Thurstone’s experiment, Wallis and Friedman concluded that “it is probably not possible to
design a satisfactory experiment for deriving indifference curves from economic stimuli applied to
human beings” . This conclusion was based upon the following –
a. Hypothetical choices do not elicit actual preferences
b. A long period of time and many different situations must be observed to obtain preferences;
and
c. Effective experimentation would require control of virtually the entire existence of the subject.
The author recognised the challenges as specified by Wallis and Friedman and also the useful
features of Thurstones as well as Rousseas and Hart experiment’s to determine a procedure of its
own .

Objective of the experiment


To determine whether the properties of the experimentally determined indifference curves meet the
properties of theoretically determined indifference curves or not -
1. Non intersection
2. General negative slope
3. Convexity to the origin

Method of Experiment

Reference point is chosen -P0 (determine the indifference curve passing through P0)
Regions :
(1) An accept region (A) - of all the points to which P0 is not preferred
(2) A reject region (R) -", of all the points to which P0 is preferred

The experiment determines the indifference curve by the boundary separating the two regions viz
Accept and Reject.

Two major difficulties:

1. What motivation is there for a person to reveal his true preferences?


a. Pay off procedures – Subject is motivated to reveal his true preference by making his
final outcome contigent on the indifference curve he has drawn.
2. How can a person be expected to consider every, or even many, points in the commodity
space?
a. Complexity reduction procedures – Can be used where saturation occurrence point of
commodities is known.

Summary of Experiment

1. Seven subjects were given training to draw indifference curve. The subjects easily
understood the procedure given their educational background i.e majors in economics,
mathematics or engineering.
2. First experiment was conducted with Pen and Money bundle on day one and second
experiment was conducted on the following day with Pastries and Money bundle. The
subjects drew the curves basis the choices made for commodity bundles offered to them.
3. The Indifference curves drawn on the separate graphs by the subjects were very consistent
when superimposed.
4. Of the 21 consistency checks given to each subject, and chosen to be difficult (i.e., very close
to an indifference line), an average of less than one choice per subject was actually in-
consistent with the appropriate indifference curve.
5. None of the subjects seemed to display any post-payoff dissonance when the actual payoffs
were made.
6. None of the usual properties of indifference curves was forced upon the subjects.
7. The authors suggested the following to improve future experiments –
a. A larger sample size.
b. More no of subjects with variety of commodities in different periods of time.
c. Validation of the curves drawn – Sensitivity of indifference curve with respect to a
reference point ,through a barter situation amongst individuals and prediction of
choices were suggested for validation of the experiment.

Learnings/Feedback

1. Through the paper we could understand that the empirical determination of indifference curve
is useful for determining preferences in non-market conditions apart from determining
consumer demand.
2. The paper also proposes an easier empirical method to draw indifference curve which
hitherto would have been difficult to obtain from the actual expenditure data.
3. Indifference curve gets affected by change in the budget/income levels, which was not
considered in the experiment.
4. The experiment considered commodity bundles of pens and pastries upto 180 and 15
respectively. The no of commodities in a bundle can be rationalised to reflect the actual
requirements, say +/- 50% of per capita consumption to guage the actual consumption
preferences.

You might also like