hierarchy from exchange joseph m. whitmeyer unc charlotte august, 2003
TRANSCRIPT
![Page 1: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/1.jpg)
HIERARCHY FROM
EXCHANGE
Joseph M. Whitmeyer
UNC Charlotte
August, 2003
![Page 2: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/2.jpg)
Hierarchy is ubiquitous
Anthropological evidence on human societies (D. Brown)
Discussion Groups (Bales) Someone talks most (usually 50%) Big talker deemed to have most
influence
![Page 3: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/3.jpg)
Theories of Hierarchy Exchange of prestige for ...
(Homans, Coleman) Expectation States Theory (Berger,
Ridgeway, Webster) In process of social learning
(Henrich and Gil-White) Reciprocal influence of status and
exchange (Eckel, Thye)
![Page 4: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/4.jpg)
My Study Task-focused group
Achievement (get prize) Choice (choose between options) Both (win, but with your candidate)
Only concerns selection of leader Computer simulation
![Page 5: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/5.jpg)
3 Assumptions Human behavior: consistent,
similar, motivated Humans value prestige (Veblen) Leaders provide collective benefit
to group
![Page 6: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/6.jpg)
The Simulation Model Each iteration each member
decides whom to support Support member giving highest
payoff for YOU Know how much support each
member had in previous iteration
![Page 7: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/7.jpg)
Payoff components Perceived ability of member Amount of support member has Value of collective benefit Rivalness of collective benefit If supporting self, value of prestige
![Page 8: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/8.jpg)
Parameters and variables 1
Group size n Collective benefit: B > 0 Rivalness: 0 λ 1 Individual share:
S = [(n - n + )/n]B Net value of prestige
Normally distributed Usually mean > 0, s.d. = 1/2 mean
![Page 9: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/9.jpg)
Parameters and variables 2
Ability to deliver B: 0 Ai 1 Spread of ability perception
0 U 1 U = 0: complete agreement about
abilities U = 1: maximal divergence For choice, spread includes
heterogeneity of values
![Page 10: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/10.jpg)
0
0.2
0.4
0.6
0.8
1
1.2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
SPREAD OF ABILITY PERCEPTION
CO
RR
ELA
TIO
N
n = 10
n = 30
n = 50
Figure 1. Correlation Between Actual and Perceived Abilityas Function of U, for 3 Group Sizes.
![Page 11: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/11.jpg)
Parameters and variables 3
Production function for leader effectiveness linear accelerating decelerating S-shaped (logistic) inverse S-shaped
![Page 12: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/12.jpg)
0
0.2
0.4
0.6
0.8
1
1.2
0 0.2 0.4 0.6 0.8 1 1.2
PROPORTION BACKING LEADER
PR
OP
OR
TIO
N O
F B
EN
EFI
T
decel
linear
accel
logistic
inv. logistic
Figure 2. Production Functions for Leader Effectiveness.
![Page 13: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/13.jpg)
Equations Let sj be j’s share of support, Pij be i’s
perception of j’s ability i’s perception of j’s effectiveness:
eij = sja Pij
a = 1: linear a > 1: accelerating a < 1: decelerating
S: eij = Pij / (1 + exp[10(.5 - sj)])
inverse S: eij = {.5 - .1ln([1/sj] - 1)}Pij
![Page 14: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/14.jpg)
Equations (cont.)
Let uij be i’s net benefit from supporting j
Let vi be the net value of prestige for i
uij = eij S for i j
uij = eij S + vi for i = j
![Page 15: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/15.jpg)
Results - Consensus
Consensus is quick or not at all 1-6 iterations for n = 3 to 10 More iterations for larger groups
Consensus is stable: no one ever switches from leader
Like discussion groups
![Page 16: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/16.jpg)
Results - By p.f.
To standardize magnitude of results, collective benefits (Bs): 70 for decelerating 95 for linear, inverse S-shaped 395 for S-shaped 495 for accelerating
U set at 0.2. Ability spread evenly from .8 to .2.
![Page 17: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/17.jpg)
0
0.2
0.4
0.6
0.81
3 6 912 15 18 21 24 27 30 33 36 39 42 45 48
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
PROPORTION FULL
RIVALNESSGROUP SIZE
Figure 3. Production Function: Linear
![Page 18: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/18.jpg)
0
0.2
0.4
0.6
0.81
3 6 9
12
15
18
21
24
27
30
33
36
39
42
45
48
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
PROPORTION FULL
RIVALNESSGROUP SIZE
Figure 4. Production Function: Accelerating
![Page 19: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/19.jpg)
00.
20.
40.
60.
81
3 6 9
12 15 18 21 24 27 30 33 36 39 42 45 48
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
PROPORTION FULL
RIVALNESSGROUP SIZE
Figure 5. Production Function: Decelerating
![Page 20: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/20.jpg)
00.
20.
40.
60.
81
3 7
11 15 19 23 27 31 35 39 43 47
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
PROPORTION FULL
RIVALNESS
GROUP SIZE
Figure 6. Production Function: S-shaped
![Page 21: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/21.jpg)
00.
20.
40.
60.
81
3 7
11 15 19 23 27 31 35 39 43 47
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
PROPORTION FULL
RIVALNESSGROUP SIZE
Figure 7. Production Function: Inverse S-shaped
![Page 22: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/22.jpg)
Group size & rivalness:
For all p.f.s, proportion reaching consensus is lowest with high group size and low rivalness
With n < 6, rivalness 0, all reach consensus
With n > 25, rivalness 1, none reach consensus
![Page 23: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/23.jpg)
Interactions:
Rivalness 0: For linear, decelerating, inverse S-
shaped, increase in group size has NO effect
For accelerating, S-shaped, increase in group size means consensus goes to 0
![Page 24: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/24.jpg)
Interactions (cont.) For linear, accelerating, and S-
shaped: negative effect of rivalness on proportion reaching consensus increases as group size increases.
For decelerating and inverse S: group size does not affect negative
effect of rivalness group size has little direct effect on
proportion reaching consensus.
![Page 25: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/25.jpg)
Implications
Crucial period is when support low Since
prestige is awarded consensually in large groups
and prestige is awarded automatically, then
prestige is awarded especially for nonrival goods.
![Page 26: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/26.jpg)
Spread of ability perception
U has strong negative effect on proportion of groups that reach consensus on leader
Its effects become stronger as group size increases.
So: to reach consensus, some agreement among group members about relative ability is necessary.
![Page 27: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/27.jpg)
00.20.40.60.81
3 7
11 15 19 23 27 31 35 39 43 47
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
CONSENSUS
SPREAD
GROUP SIZE
Figure 8. Effect of Spread of Ability Perception (U) and Group Size.Rivalness = 0; Decelerating P. F.
![Page 28: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/28.jpg)
Mean value of prestige
Question: Suppose two populations differ in mean value of prestige. Will leaders tend to come from one?
Answer: No.
![Page 29: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/29.jpg)
Mean val. of prestige (cont.)
Suppose As have higher mean than Bs, s.d.s same (5). Then: Chances of random A having
higher preference for prestige than random B are greater than the chances that an A will be leader.
Difference of 1: .556 vs. < .52 Difference of 5: .76 vs. < .57
![Page 30: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/30.jpg)
Summary
1. Stability is reached quickly.
2. Full consensus is permanent.
3. Increasing spread of ability perception diminishes proportion of groups reaching consensus.
4. Production function probably decelerating or inverse S-shaped, perhaps linear.
![Page 31: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/31.jpg)
Summary (cont.)5. Critical period for deciding group leader is initial phases of
gathering support.
6. Prestige awarded mostly for nonrival or nearly nonrival benefits.
7. Populations with different mean net values of prestige produce similar numbers of leaders.
![Page 32: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/32.jpg)
Next Steps Items 1 - 3 parallel experimental
findings Items 4 - 7 are predictions to be
tested
![Page 33: HIERARCHY FROM EXCHANGE Joseph M. Whitmeyer UNC Charlotte August, 2003](https://reader034.vdocuments.site/reader034/viewer/2022051621/56649f575503460f94c7c2b1/html5/thumbnails/33.jpg)
On nonrival benefits (#6): Exs.: Leadership, entertainment,
protection or defense, and information of some types.
Suggests theory can be applied to socio-emotional groups too.
Apply to different historical periods, with different non-rival benefits.