Condorcet's Jury Theorem

You might also like

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

The Interactive Conceit

by Manu Panizo
Mathematical proof of Condorcets Jury Theorem
This can be proven mathematically:
Consider three1 voters, each with a competence of 0.51, which means that each of
the voters has a probability of 0.49 of being incorrect. We want to know the probability of
the majority being correct.
Number of voters: n = 3,
Competence: p = 0.51,
Number of voters that make a majority: m =

n +1 3 +1
=
= 2.
2
2

The probability of at least a bare majority being correct is the probability of m


being correct, , or exactly n being correct. In our
being correct, m+1 being correct, m+2
case, the probability of at least a bare majority being correct is the probability of two voters
being correct or three voters being correct.
This can happen in three2 ways:
1)

A and B are correct

C is incorrect

2)

B and C are correct

A is incorrect

3)

C and A are correct

B is incorrect

We use an odd number for simplicity, but the same argument works for even numbers by adding a tiebreaking rule.

The formula for choosing r elements from a group of n is

n! . In our case:
r!(n r)!

n
n!
3!
=
= 3.
=
r
r!
n

r
!
2!1!
(
)

Because we only need one of these to happen and we do not care which one, what
we will measure is the probability of any of the possible two-element combinations taking
place. Since each combination is equally likely, the probability of two being correct will be
the number of possible combinations times the probability of one happening.
Take, for example, the first case:
P (A being correct) = 0.51
P (B being correct) = 0.51
P (C being incorrect) = 1-p = 1 0.51 = 0.49
The probability of these three happening is:
P (1) = p p (1 p) = 0.51 0.51 0.49 = 0.1275 .

And the same for the other two possible combinations:

P (2) = 0.1275 and P (3) = 0.1275 .

Because all voters are equally competent, the probability of two of them being

correct is:

3
p p (1 p) = 3 0.1275 = 0.3825 .
2
In general, the probability of h voters being correct is:

n
n h
h
p (1 p)
h
And the probability of at least m voters being correct is:

n
n
h ph (1 p)n h
h =m
In our example, the probability of at least two voters being correct is:

3
2
1 3
3
0
(0.51) (0.49) + (0.51) (0.49) = 3 0.1275 +1 0.1327 = 0.5152
2
3
As predicted, the collective decision of three voters is more likely to be correct than

the decision of one single person (0.5152 is higher than 0.51). And the probability of
making a correct decision will be higher as the number of voters increases, tending to 1 as
the number of voters increases to infinite3.
n

n
n h
lim p h (1 p) = 1
h
h
h =m

Manuel Panizo, All Rights Reserved

www.theinteractiveconceit.com

By definition, the sum of the different probabilities of all possible outcomes will be one, which is precisely
the case of the number of voters tending to infinite.

You might also like