OF • MO DE l.

S
TH f. ME TA Pli OR JC AL USE
th m~ els ilnd the ir ma r.y v:i ria nis arc diverse rep res ent ati on s of
Th ese ree col lec tiv e act ion . Mu ch mo re wond rk
r an d stt ll-e vol vin g the ory of
a ~ro ade
l be nee ded to dev elo p the rhec•ry of col lec tiv e act ion int o' a r·eliable a
Con~idcrahlc pro gre ss has bee n ma~e
wd
use ~I fou nd ati on for pol icy ana lys is.
thr ee dec ade s by the ori sts and em pir ica lly .or ien ted so aa l
d~nn~ rhe P~
fll'St ,·aria'nts of this theory have
scientists. Th e ~·ccp1ng co,:1clusions of th~ re
wledge involving .many mo
giv~n way to a more qualifiCd body of kno : ·
variables and explicit base C(,nditions. :· agr eem ent
, the ory , at pro'-7okc-s dis
As a!1 evo lvi ng . r.1rher th~ ~ com ple ted of son ,c var iab les an d ho w bes t
regarding rhe im po rta nc e or: ins ig~
fficance
rk, ·partic-
from ,nore ~cccnt- wo
to specify kty tt14rionships~~ The ·res11lrsca s of relcvatjc en, pir ica l se tti n~
ularly wo rk foc usi ng on rh~·dynamic asp c pre dic no ns t~a n did ear lie r mo d-
ro ge ne rat e m~ ·re opt :.m isti
~v e be gu n (19 81 , 1~ 84 ):a nd krc ps and
rtic ula r, the wc ,rk of Ax elr od
,els; see, in pa
n ( J 98 2). Th is is on e pf the :no st exc itin g areas i~ .th~ soc ial sciences,
Wi lso y o¢~urrcd, some de ep
h con sid e'ra blc· ~u mu bri on has alread
(or althoug
s rem ain un ans we red . So me of these [Link]( arc key to under-
question
how ind ivi du als ;oi ntl y using o co mm on-pool: resource ·.might ~
sra.1W 11g g their own •
an cff~ctivc form of governing· and flla n:igin
able to achieve 2~ . ! .
d in 'Chapter
commons. Th ese puzzles arc examine nmon•pC:,ol •r~sc,u rces, however 1 : •
Mu ch r ha t has heen written abour coi
tic all y acc ep ted rhc ear lie r mo de ls and ·c_h~~- presumption of a •
h&S uncri
seless tra gedy tN eb cl 19 87 ). Sc holar s have gone .~so far as to rccom•
remc;,r be. required rcad-
: mend that -Hardin's 'Tragedy of the (:ommons' shoul~ n bcings."s Policy
for a~I hu~a
in·g for all srudenrs ... afld, if I had my way, thr~e origi~al
pr ~p rio ns . haV e· relied to a l~ ge cXte,it on one of the
bur- those 2n cm pri ng to use the se mo dels as_·,hc basis for policy
models, :;a mctap~orical use
quently have ach i~vcd little rriore than
prescription fre .l
of rhe models. ••
arc used a, merap ho rs, an autho r us~ ally points t<1 the
When n1odels d one or two
one or rw o varia bles in a natural setting an
similarity hcrween
,
-- --
----..-----
• ·-- --

7i fl' I' 't, I I 7'0

ff l
r~
 •
Goven,ing the commo,,s
il... . . •.
iables 1n a mod~I . If•[Link] attealntion ro. sim -r1t1cls 1s all tha ~ .1s .intended
h .. rap'd
rpose of • y convcy,np • f
var rves t c usu pu
by rhc n1crap orf, •~•~cTh
h' rhrec m
od els have frequen t! bee o in nrrna,
.•tion in grap IC or~. es e
r anothe r pu rpose. The si milarity~ n used n1cta•
all)', ho\Vever , ~o the many
• phoric
ing a reso urce in 3 na tu ral sc...: twaneednthe
individuals joindy ,- ~s , b . •. . "-A ng
: d • nhsu ~pt_•,m~I, r~sult in the model h• • rnhcany
. . I • •intIy·!~ ro tuhact1nrg en
ind1v1dua s Jo . un er s1m 1 a,1ncs· are prese 8 ferring to Uo1.

u,cd to convey a ~J ragedies of the commO~ " "coll nt:, Yacre_aon pr0b
\IC

, mnural scetings as
" u • '
lems, pru;oner s " h b
~t
d~:•I em m as ,"
se
..
rv
op
er
en -acc es s •so.'urces "ectJve- "
re
frequcndy [Link] , even
. es to inv k .
or co m m on
~
;
propcrtY resources,\ r e ht in a,n incxorablCiproccss of ; e an ~mnge ~f
o

' hdplns individuals caug in rhe [Link] lO 1988 , 1•.-.UCes~r

own rcsou~ces. An. '.a •
niclc
ries can
• '
:he nian agcd
C!tl
O
fo~ngCttheir
, nt
full on-
r 0 [Link]
' ro as se n th at fishe 5
0 ,nist goes
so (ar a~ n de vices fisherman wuc,.clless
izcd thar .'61c:.h to their ow •·' overexp1
if iris rc(~gn
must hnvc cffcaive hegemony
'- sto ck s,"
over thcn1.
.
an ~ ato avoid disaster~ manag
n~ more: th30
ers

:~
:
voke grim images by briefly
Pu~lic offic:ials someti?1cs Jo as self.
~ed versionsof the model" presuming,
alluding to the po,pulanoc [Link] occu r in all narural semngs. The Cana the
dian

~r~
evident, that rhc sam e pr
ries and oc ean$ , for ex am ple, c3ptured the color of
minister of fithe : : '
models· in a 1980 speech ries. with everybody
fishe
you lcr loose that ki nd of N:onqmic sclr-,ntertSt an ro no individual, you end
If 'fron1 .1 ·rpourcc rhar bclo
np
bad
fishing as he wants, tilkingour mnd ypursclf. In lrcc fishcric, 1 good times cregrc le$$
~ up destroying your neighb more ~ ts to cha.~ fewer :and fewer fi,h, producing
rimes, anraaing more ind g ~ re anJ more people. annivcrwy meeting
~ les s m oney to ds-vidc am on
and rhc 50th
·(Romeo LcBlanc, speaking at liishermcn, Morc;h 19, I'iRO;
~ of rhc UnneJ •Maritime atthew, ind Phyne 1988)
quoted by M
~
se , th at C an ad ia n fisheries universally mer that
The implifrion, of cour
W ai
in co rre ct infe re nc e.• Bu t ma~~ ob~crvcrs have
-.~
C

descriptio~ - an empirical ly
ce s W 'e lik e' those specif1ed 1n rhc three
m os t re so ur
I~ come to ~ um e thJt su m ed th at the in di viduals have been caught
been as
models. As such, [Link]!i ns have had an equally
I CJ in a grim trap. The resulting
po licy rc co m m en da rio
~ 1n_t (eJ1fa
~ ... 12- 1 I3
°
grim character. ••
I. ~ J~ /( ir R R E N T PO
LICY PRESCR IPTIO~S -~
~
y
Leviathan .as the "only'' ,ua
'~
I ,,
v '~'tM£
•
• ....

ed , for cxamp c, t at
rragedy
I h "bee usecdoftHtheroug
be :olv h co·
Ophuls (1973, p. 228) argu tal problems cannor . h ajor coercive
en
of rhe commons, environm nale for government wu m
'~
I ., operation .. . and the ratio
8
I ,.
 Refl ei,io ns OIi the co,,, 11,o ns .
pow en is over whe lmin g. ....Qph als conc :l~d
cJ th~t ·•e~~n if w~ ~voi d t~e.
rragedr o( th! com mon s, ii will on/'1 l,e- by reco
urse .,0•'1he trag_,c _ncCCSSlt}
Gar re~ Htlr d,n nr~u ed a • ...
ol Lev, nrha n (197 3,·p . 229 ; emp hasi ~ adde d).'·
~ a .. cloud of agnor- ..
decade ahe r his earlier anicle that we: are enveloped
1ystem1 and the : •
a~ " •hour ..thr rrw narure o~ the fundamental polit·i~al
!fleet o( c~ch on rheprescrvation of chc cllviron ment".(1978, p.31_0). The
ent him &pm pre1um1nt1 rh!t
~lou d _of •11nor11,!ce did not, how ever , prev ffl4 were '. wha t he calle d a
: the only alre mat ,ves to the com mon s dilem
--sociatism," on the othe r
priv ate enr~rpf'.'&M' 1y1r em," un the onr hand nt
d thar •rhe alcernative
(197 8, .P• l l ◄ ). ~ith ·thc· RKsurance of one· c~nvince
is too horr i(yin 1 to cont emp late " (196 __, p. 1,2◄ 7),
Har din
~f~ cco mm ons
- •whatever force
illdicared rhat change woulJ hav.~ to be instituted with
' ~1 be ~equired to ~•k ~ the change: [Link]" (l 9?R
, p. l f◄). In othe r v.:ords~
e be respo~s,ve to
ii nain .•• to be avoided 1n " crowded wor ld, peop le mus viathan,' to use
~
a .•Le ~~
• coe rav e force outside their individual psyches,
• .•
Hobbea•s term " (Hardin 1978, p. J 14 ).
d trag-
The presumption rhat 311 external [Link]·an 1s ncc·cssary ro avoi
ernments
edies of rhc commons leads to recomn,enclarions that .c_entral gov ed thar
ner ( 1974) opin
conn -ol mos t •natu ral reso urce syst ems . :He1lbro
, wou ld be necessary to
-UOn gov crnm cnrs , .. perh aps mili tary gov ernm ents
onian view, Ehren-
achic¥e control over ecological problems. In a le~ drac· (an not be exp ecte d
feld (J97 2, p. 322 ) suggested that . if .. p~ivate interests
ro pro rect the public domain then external rcgularion
·)>y public agencies,
ao,ernme,rs, or. international authorities is needed.~clev In. an analy$is of. the·
pro blem s invo lved in warer resourc~ [Link]
CMrurhers and Sro ncr (Jj8 l, p. 29) arg~ed ~at wi1h
in clopini cou ntri es,
ouc. public· control, ~
•o•[Link] and soiJ erosion of communal pasrures, or less
AYCGIIC cosr," would result, Th~y ,~c111cluded
~sh at higher
that "co~n,on prppcrty: r~- ..
aources rtqui,e public control if economic cfficicncy is to rrsult fron, their
~e t~ cen-
..
devtJopmenr" (1981, p. 29; emphasis added). The p01icy advi
1

tralize the control and regulation of natural rcso~rccs, such

as [Link]
ely, paiticularly in
..
lands, forests, and fisheries, has been fc1llowed rxt~nsiv
Third World countries. , :
zed control is
•
One way ro illusrrate ·the,e proponent~• image of ·ce·~trali
ro modify the Hardin her,ler gar.1e u5ins t~,e [Link]~~s tl,at
u~dcrlie this
t .A external
•4
policy advice. The proponents of centralized c:ontrol :wan t the central
eg, tha
pem mr nc agency 10 decide rhe ,pcc:ific _berding s~t y *ill decide I
aucborky con1idcn be• (or .the [Link]~n: The central authoricanimals can
w,ocan use the meadow, when d1ey can use it, and how m~ny
impose.a
be grazed. Let us [Link] [Link] the central authority decides to
authority to
penajl}' of 2 profit unitS on anyone who is considered by that
9

--.--
t;;;;; :t,:j
 e commons
Governing th J
,J s su m ing char tlie kno ws the
t srrate 8 )', A en ~ ~ ,. aglgency
IJ be using a· de/ec of the mead~w ( l) and ,.~3n~un,a, ,n r ~u!iscover and
1· sustainable yhieldd usm . g th e def.e a strategy the IGy restructured B~e
r e r /~ w h
,, J ~ i u any e central aurhor~ry ;, repre~n in ertamree 2.• Now, t e
te
imposed hrGy the 2··· , C Q operate). Both
pla
ce,ve 10 profit
J .on to amc •,s ,coop erate ~ ld ha v Y • .
soluti an the zero [Link] rhev wou rhe Qe re~e,ved in Game J.
~ th e,
units each, rather rh oriry accura r~Jy d e r e r ~ n
c irv ,:aco~nry a ~0 nunon-
t of tons and
ex_ rerna l au ig n s rhis ca p a , o rs aft '
.:, I f an isuo usly S JS -~ ,

o ol -rcsoW 'ce, una~·b ,! _ _

1 ce , th
d
en a centralize y•~~n~ c u ~ -
• m pu a J
P • - cno:ns nonco an optimalJ err1c1enr enqguda1nbd-
~ unfailing1y san in ~ tr d c r game ro gen. [Link] . e cost o f cre
ati
a rd .d e n th
e pr oblc:. m and
e H I g Jv to
:J form th It h d . L.1tt e co!hs•. ~ration 1 s
g e n o us ro th
. ~h e~s. en as ex o
rJu?1 [~ r ~re suce a,:a agency. • 1sf ,sGsemc 2 .' i

ma [Link]•~g ro a
. 1<:1u uJ c_.u.J as a p a r a m e te vic'e to ce nrraJize
foUowing the .ad the accuracy of
•
;J
is n o r 11 b
achlevCd y
~
he op ~ U JJ al c q u ~ljbrium u inp c ions concerning nd zero costJ
T on a ss
ver, JS based g reliabiliry, a .al agency
,'J control, howe ie s, sa n ctio nin
ring apabilit reliable information, a cenrr or rhe
a tif n , m o n ic o
'# Ulform n . W
d
ithour valid an setting the carryins capaciry nor
is tra tio
of admin _l e rrors, incl1,1dins w ho cooperare, or a• '
e s~ v~ra erd ers
,"J ,ould mak o r too lo w, santtionin
s h
f incomplere in
form
aJI forms o the
fine too high f~Ors. The impliprions Omf ple, I shall focus entirely oanbour
- ~ c t i o n i ~ g desting. ·However, a.s a n exnacy's incomplete informarion t t~~
~ ripn a r e intererisin1·from a c e n tr a l ase Umprion o f Gam~ 2. is cha s
impJicarions
a
g j~ . T he implicit a$5 rd ers co srless J.r and impose
r~ e
- the herders'
stra
o ~ ir o rs aJI a ctions o f rhe h •
a gen cy_ m
cen tr al
ec rl)'; . cy h as (o m p lete ~lorma~on
~[Link] co
rr
fflC rh a r rhe ccnrral agen om p le ce inlormatJon
- a$ $u r in c
In Game 3, we ~paciry of the mradow, bu central agency conse•
ying/ The
i ..• about rhe carnr icuJja ·actions o f rhe herders. .
••abour the pa .. '

~
0
• -2
11 ,v
-lv-f. .•
:
_, g '
' .. I

~
·,
'
1·0 _, -2
9
'0 •
e1,&u·~ ..

1Vll~ {/°µM_

f It n'non.
cowpklt m/ort
I
I
.s..
• - .. -
U IJ J, C lf lU •
a....
,,,....- · .
-AUfuurny
(; M n t l: Tht ctnEriJ
Fi,ure I.2.
 ~
' Re{leetions on the commons
5
quently makes errors in imposing punishments. Let ~ as~ume thai .t~c
·~
~
central agency punishes defections (the corre~ ~esponse) with ~robabaltty
1 and fail, to punish defections Y.-ith probabthry 1 - ~ (the errone~us ~
response). Ler us also assume rhar the cenrral agency punishes coopcrat!ve
acdon.s (the er-raneous r~sponsc) [Link] probab~lity ~ and. ~ocs not punish ~
cooperative actions (rhe cocrcct response) wirh probab1hty t 1- x. The
I . . ~
payoff parameters are illuStrared in Figure 1.3.
A central agency with complete information w~uld make no errors an its ~
punishment level; in that case, x = 0 and y = 1. Game 2 w':lu!d then be
a spcciaJ case of Game 3 in which x = 0 and y • 1. However, 1£. the central
~
•gency docs not have complcrc information about the actions of the herd- 4
ers, it imposes both types of sanctions correctly with a probability of 0.7
(x~ 0.3, y = 0.7). An example of the specific payoffs for this pmc is 4
1hown as Came 4 in Figure l.4. Given rtiis payoff smicrurc, rhc herders
apin face a pri,oncr•s dilemma game.. They will defect (overgraze) rother '"=-~
than cooperate (graze within the carrying capacity). ·1n Game 4, as •in the ~ ~
original Game t, the equilibrium outcomes for the herders were (0, 0). ln
a game in which a ccnrraJ agency sanctions correctly with a probability of
~~
0.7, the equiJibrjum outcomes arc (-1.6, -1.6). The ·equilibrium of the -~"~

regulated game has a lower ~~lue than thar of the unr~gulate d game. Given
the carrying capacity and profir possibilities of Game 1, the central agency ~ ~
m_usr have sufficient inform~io,:a so that it can corre~ly [Link] ctions
wrth a probability grearer t~an O.?S to avoid [Link] the herders to the
(D, D) equilibrium. •0 ...•.. , ;.
. •
~-~
.
. •. f'(U rJi OYI ;~lt~ OJJ w1,f~6>' ,
•
. '.'
I
f

'
. .. ~
~
'

...
' ..
..
\.

v,\ .... . . .'

. ...
~
'

IO• b • I• ZX I 1 •·21

eltf ~ofn I ~- ~ I I • ZV • I • 2x
. .
.
,ttuJ a~~ I I

'
• '

'

.. '
'
.
.'
• ' '

.' ~
f
.'.
~
.

•
.
. Fiprt I.J. Gamc-3: The ccnrul-iurhoricy pmc with ,ncoml)ictc 1nforfflltion.

11
. .
•
 I

, .
. 0
G ov e, ni ng th e com,,..ons
•
;
i'
, , . .
' ,

94 1,1 6 I I t 6. I • I• I 4
I . 1. I ,4
,~
,/

I I
1
'9.6 I I - t.6 :
- . '
.
,
..
□
,
•
-...... -. - -
-----~- ~ , '

. ,
figure 1.4. Game .-4: An e,xlfflple of rhe ccntr
al•authoriri ;.me with incomplete information
~
.
c) \~ ~ ~ tu A M W -
u --- ~ 11,vJi,. ·~-- ru .rr
iu ti11 x ()