Polsci 281: Path Dependence

Front 1

path dependence

Back 1

essentially: "history matters"

Front 2

increasing returns

Back 2

more begets more
same as postive feedback

Front 3

self reinforcing

Back 3

same as feedback looks
existence creates other entitites that reinforce it
one could not exist without the other

Front 4

e.g) Ann Arbor v. Jackson

Back 4

cities went on different paths

Front 5

Ann Arbor

Back 5

large public university, then it becomes a railroad hub and grows exponentially
current population: 120,000
Median Household income 46,000
5% live below poverty line
2.8 crime

Front 6

Jackson

Back 6

biggest 4-walled prison
birthplace of the Republican party
current pop: 36,000
Median household income: 32,000
15% of population below poverty line

Front 7

Chaos/ ESTIC

Back 7

Extreme Sensitivity to Initial Conditions
--if p and x' differ by even a tiny amount, after many iterations of the outcome function, they can differ by artibtrary amounts

**small initial differences get amplified

Front 8

Basic Urn model

Back 8

Urn contains balls of various colors: the outcome equals the color of balls selected

Front 9

Bernoulli Urn

Back 9

Assumptions
U= {B Blue, R red}
select and return ball

P(red)= R/(R+B)
outcomes independent

Front 10

Polya urn

Back 10

Assumptions
U= {1 Blue, 1 Red}
select and return
add new ball that isthe same coloar as the ball selected

increasing returns: when you pick a blue, you are more likely to get blue in the future

any history of B blue an R red is equally likely: if RBBB or BBBR- this is hte same thing- the order doesnt matter,, just the ratio of balls do

Front 11

State Dependent

Back 11

if outcome probability depend on some finite set of states

Front 12

path dependence

Back 12

if outcomes probability depends on the sequence of outcomes
in this case order matters

Front 13

set dependent

Back 13

outcome probabilities depend on the seto f outcomes but not their order
e.g) last semseter took 4 classes. doesnt matter which order i took them in.

Front 14

balancing urn

Back 14

Assumptions
U= {1 Blue, 1 Red}
select and return, then add ball of opposite color

it is sets of negative feebacks and leads to equilibrium distributions

Front 15

fading past urn

Back 15

U= {1R, I B}
select and return add 2^t
select Blue then add 2^1 blue
next time select blue then do 2^2 blue

very strong path dependence
no equillibrium

Front 16

sway of the past model

Back 16

every single one is dependent on the exact last one
it is a set of equillibria
path dependent
** p0ast takes on more weight over time
early choices matter more than current ones

Front 17

externalities

Back 17

when activity causes external costs or external benefits
possible positives: choc + peanut butter = reeses
neg: air pollution

Front 18

QWERTY
positive and negative externalties

Back 18

QWERTY has postive externalities on other QWERTYs
negative externalities on DJKLIEFS
real example of path dependence
QWERTY came out first
made consumer learn QQWERTY
once learn i dont want to try another one

therefore making negative externalities on other keyboard companies because have different combination of letters

becomes a complex system because it bumps into other keyboard companies