Cooperation on the tundra: Field experiments with Saami herders

Cooperation on the tundra:Field experiments with Saami

herders

Matthew Gwynfryn Thomas

Human Evolutionary Ecology Group

University College London

11th May 2015

“The essence of dramatic tragedy is not unhappiness. It resides in the solemnity of the remorseless working of things.”

Private goodsClothes, cars, food (sometimes)

Common goodsFish, timber, coal, planets

Public goodsAir, streetlights, NHS

Club goodsCinemas, religions, the EU

{Subtractable

Excludable{

Photo © David W Lloyd

West et al. (2007)

West et al. (2007)

Kin selection Direct reciprocity

rb > c

Tit

Tat

Kin selection versus reciprocity?

The relative effects of reciprocity (RA), kin selection (KS) and tolerated scrounging (TS) on food sharing

Tag-based cooperation?

Tit for tat?

Free-riding?Policing?Reputation?

rb > c?

Study aims

1. To understand the dynamics of cooperative behaviours within and across herding groups

2. To investigate biases towards in-group (and other groups)

3. To investigate how kinship and group membership affect cooperation

Outline of the rest of this talk

1. Overview of Saami pastoralism

2. Public goods games

3. Gift game

4. Social relations model of gifts (exploratory analysis)

Study site: Finnmark, Norway

~2,400 Saami reindeer herders

378 license owners

(75 in study district)

Households form cooperative groups: the siida

Summer siidas contain from 10 to 150+ people

Summer siidas split into smaller, family-oriented winter siidas

70°

65°

60°

20°10°

100 km

From Næss (2009)

Cooperation (and conflict) on the tundra

Exchanging labour, tolerated encroachment, migration corridors

Flexible herd management strategies, based on e.g. herd sizeand the behaviours of neighbours

Herd accumulation as risk-reduction strategy

Too many reindeer: quotas enforced, but equivocal evidence for overgrazing

Conflicts, especially among winter siidas

Herd abundance over timeNæss & Bårdsen (2013)

Little trust in winter siidasHausner et al. (2012)

Field methods

Questionnaires about demography, siida membership, cooperation

Two public goods games (one shot)– giving petrol to (i) their siida, (ii) their district

Gift game (one shot)– giving petrol to other herders

Male: 61

Female: 4

Sex

Married / cohab-iting: 17

Never married:

10

Other: 3

Marital status

Primary / sec-

ondary: 18

Upper sec-

ondary: 6

Uni-versity:

6

Education

25 30 35 40 45 50 55 60 650

1

2

3

4

5

6

7

Age distribution

100 300 500 700 900 1100 1300 1500 17000

5

10

15

20

25

30

Distribution of herd sizes in 2012

Public goods games

1. Everyone gets 5 litres of petrol

Siida

2. Decide how much to donate to siida/district pot

Siida

3. Multiply and redistribute pot equally

Siida

Marginal per-capita return rate (MPCR) matters

MPCR is pot multiplier group size

If > 1, donate everything but if < 1, donate nothing –> social dilemma

MPCR tends to be 0.5 in field-based games (Gerkey 2013) but people still donate

People donate more to their siida than to their district

These people also donated 5 litres to their siidas

MPCR = 0.004but 40% of herders donated

People report behaving more cooperatively towards their siida

Less frequent More frequentOwn siida Other siidas

Increasing return rate predicts higher donations to the siida

R2 = 14.4%

One closelyrelated siida

Conclusions from public goods games

1. People cooperate more with their siida

2. Marginal per-capita return rate important: people donated more as MPCR increases– though we didn’t systematically vary the MPCR– and a single siida drives this effect

3. Researchers should play other games – common and club goods

Gift game

1. Participant endowed with 3 x 5 liters of petrol

2. Give all away to one, two or three others

Gift game predictions

1. Herders will give gifts to closer relatives

2. Herders will give gifts to members of their siida

3. Gifts to younger family members

Colours are siidas

Circles = 75 license owners(Size is no. gifts received)

Gift network

Colours are siidas

Circles = 75 license owners(Size is no. gifts received)

Filled circle = 30 interviewees

Gift network

Colours are siidas

Circles = 75 license owners(Size is no. gifts received)

Filled circle = 30 interviewees

Lines = gifts(Thickness is gift size)

"Deserves it“"Good reindeer herder“"Always empty of fuel".

71 gifts given(45 within siida)

Gift network

Gifts not given preferentially to neighbouring siidas

No spatial clustering of gifts

Giver/recipient herd sizedid not improve fit

Neither did interactionsiida relatedness

Belonging to the same siida best predicts gifts

Good herders: 8

Young/new owners: 5Current or future

reciprocity: 2

Old friend: 1

Need help: 1

Lazy: 3

Family: 2(No reason given): 1

Other siidas

Good herders: 3Young/new owners: 1

Current or future reciprocity: 12

Deserving: 2

Selfish: 1

Family: 7

(No reason given): 4

Good herders: 2

Young/new owners: 1

Current or future rec-iprocity: 9

Deserving: 1

Need help: 1(No reason given): 1

Kin

Same siida

Current or fu-ture reciprocity:

1

Family: 2

Non-kin

Gifts to kin Gifts to non-kin

to youngerherders

to olderherders

to youngerherders

to olderherders

Gifts not preferentially given to younger family

Conclusions from gift game

1. Siida membership predicts gifts better than kinship

2. Mix of direct and indirect benefits within own siida?

3. No age bias in gift giving

4. Reputation, reciprocity seem to be important

Gifts between siidas(exploratory analysis)

Social relations model

3

1

dyadic variancegiver variance

receiver variance

2 5

Closer relatedness predicts gifts between/within siidas

Kin – distant(r < 0.0078)

Kin – up to 2nd cousin(r < 0.0313)

Kin – up to 1st cousin(r < 0.125)

Kin – close family(r 0.125)

Giver – no. participants

Receiver – no. participants

Giver Receiver Dyad

Variance PartitionCoefficients

Generalised reciprocity 0.43

Dyadic reciprocity 0.75

Correlations

Conclusions from social relations model

1. Large numbers of gifts between two siidas tended to be reciprocated – playing favourites?

2. Siidas which gave more gifts tended to receive more (generalised reciprocity = 0.43)

3. Receivers are most important source of variation in gift giving

Limitations

Small sample in a small-scale society

Single district – other outcomes elsewhere?

Licensed herd owners only

No one wanted to talk to me

Costless cooperation

Next steps

Investigate reputation, reciprocity; winter siidas and land tenure

Comparative study with Tibet

Possible citizen science projects: – climate change adaptation– predation– monitoring land use

Ruth MaceMarius Warg NæssBard-Jørgen BårdsendHuman Evolutionary Ecology Group at UCL

Katharina OlsenJon Mikkel EiraThe herders of Finnmark

Thanks!