Tribe or Nation - Miguel (2004)

Question 1

ethnic diversity in SSA

Answer 1

world’s most ethno-linguistically diverse continent

14 of the world’s 15 most diverse countries are in Africa (exception is India)

Question 2

measuring ethnic diversity

Answer 2

ethno-linguistic fractionalisation (ELF) as a measure of diversity
- values between 0 and 1
- 0 for total homogeneity and 1 for complete diversity

interpreted as the probability that two randomly chosen people in the population come from random groups

Question 3

Easterly and Levine’s main finding (1997)

Answer 3

per capita economic growth is negatively related to ethnic diversity across all countries in the world

coefficient estimate on ELF is b=-0.02 and is statistically significant

Question 4

Miguel’s analysis

Answer 4

took the data, replotted it and extended it over a longer time frame

if you go from an ELF of 0 to 1, average annual economic growth is slower by 1.7 points in the data

Question 5

what does this mean for Africa vs. non-Africa or Africa vs. Asia comparisons?

Answer 5

ELF might explain a difference in African vs. Asian annual growth

1 percentage point lower per year in SSA than Asian countries possibly due to this association with ethno-linguistic fractionalisation

Question 6

social division and policy choices

Answer 6

beneficial policies may not be adopted if they disturb the status quo or the political equilibrium

Question 7

nation-building case study: Kenya vs. Tanzania

Answer 7

Miguel compares inter-ethnic cooperation in rural western Tanzania vs. rural western Kenya using local data on public goods

both have similar geography, climate, ethnic composition, colonial legacy and arbitrary national border (1890)

Question 8

post-colonial divergences in the two areas

Answer 8

public policy diverged a lot after independence from Britain in the 1960s in both countries in many different dimensions

characters and philosophies of Jomo Kenyatta and Julius Nyerere
- pro-Western, pro-market leader pursuing ethnic policies vs pan-Africa socialist leader fighting against ethnic policies

Question 9

how did policies diverge concretely?

Answer 9

in Tanzania, government promoted the use of a national language (Swahili) which was not the primary language of the majority of the population
- coherent Tanzania national language policy

political socialisation in schools in Tanzania
- language use, curriculum with a pan-African emphasis, support for the ruling part
- not a strong democratic impulse but promoted national identity and pan-African ideology

traditional tribal chiefs abolished in Tanzania
- movement towards a bureaucratic state and replacement with councils
- Kenya still has traditional tribal chiefs and elders

big push in Tanzania to have an equitable investment of public resources
- in Kenya, Kenyatta targeted resources to his home area and areas of his ethnic group

Question 10

Busia, Kenya vs. Meatu, Tanzania

Answer 10

busia
- average ELF=0.23, range 0-0.6

meatu
- average ELF=0.13, range 0-0.6

similarities but not an exactly matched sample
- similar livestock, cash crops, well maintenance, etc.
- busia has slightly better school infrastructure and is somewhat richer

Question 11

outcomes in busia

Answer 11

ELF and amount raised per pupil
- huge variation, some areas raise 0 while others raise close to 10USD
- downwards-sloping relationship so areas with higher ethnic diversity raise less money

ELF and amount of desks per pupil
- downwards-sloping relationship

Question 12

outcomes in meatu

Answer 12

ELF and amount raised per pupil
- range is similar to Kenya
- no relationship in Tanzania where ethnic divisions were downplayed and national identity promoted

ELF and amount of desks per pupil
- somewhat upwards-sloping

Question 13

qualitative evidence

Answer 13

Busia
- rivalry between groups over control of school committee
- social pressure rarely applied on free riders in ethnically diverse communities

Meatu
- strong national identity, “we are all Tanzanians”

Question 14

estimation issues

Answer 14

small sample size
- concerned about external validity

inability to estimate the impact of distinct Tanzanian policies

no panel data so DD approach not possible

Question 15

what does the Tanzania case show us?

Answer 15

sheds light on the origins of institutions, social capital and peace
- cultural norms can be shaped by charismatic leaders like Nyerere

post-conflict periods or democratic transitions may be particularly promising windows of opportunity