Special Feature

The Welfare Economics of Amartya Sen ; A Note

P. R. Brahmananda

A m artya S en, cu rren tly the M a ster o f T rin ity C o lle g e , C a m b rid g e
U niversity, has been aw arded the N obel Prize in E conom ics for 1998 for his
contributions to w elfare econom ics. Sen is the first Asian to achieve this distinction
of the N obel aw ard in econom ics. He studied at Presidency C o lleg e, C alcutta, and
later at the Trinity C ollege. Cam bridge. He has w orked at Jadavpur U niversity, Delhi
School o f Econom ics, w here he was inducted by Dr. V.K.R.V. Rao and Dr. K. N. Raj,
the L ondon School of E conom ics, O xford U niversity, and the H arv ard U niversity,
w here he was a professor both in the faculties o f econom ics and o f philosophy. He
has heen the P resident o f the Indian E conom ic A ssociation, the International
E co n o m ic A sso ciatio n , the A m erican E co n o m ic A sso ciatio n , and th e W orld
E conom etric Society. He has been h o n oured w ith a num ber o f h onorary doctorates
as w ell as renow ned lecturerships.
The N obel C itation refers to a) S en’s contributions to social cho ice theory,
b) S en 's efforts to link inequality m easures w ith poverty m easu res and c) S en ’s
w ork in developm ent econom ics, especially in the analysis o f the relation betw een
poverty and famines and to his concepts of entitlem ents and capability developm ent.
S en ’s w ork in these areas is vast. H e is an expert in the axiom atic approach in the
language of sym bolic logic. It is difficult to label his ideas as those of any particular
school o f political econom y. H ow ever, he should be credited for having established
his o w n special p ara d ig m in w elfa re ec o n o m ic s esp ecially in its relatio n to
d evelopm ent. He has pul forth several new concepts, som e o f w hich have been
sought to be em pirically corresponded. O ne o f his m ost am bitious undertakings is
to induct capabtliiarianism as a norm ative basis fo r political ec o n o m y in place of
utilitarianism . If he succeeds his w ork w ould h ave a perm anent and positive impact
in econom ics. In w h at follow s wc look at aspects o f his co n trib u tio n s associated
w ith the N obel aw ard. We begin by constructing the com ponents o f the A m artya
Sen paradigm , w hich is still in process.

The Sen Paradigm

(1) S ocial choice in the Pareto, B ergson, S am uelson tradition, culm inating in the
social w elfare function, deals w ith the aggregation o f choices on alternative social

Journal o f Social and Economic Development

Vol. H, No.L Jan J u n e 1999, pp. 142-153
 The Welfare Economics o f Amartya Sen 143

states o f individuals, each concerned p rim arily w ith h is ow n w ellb ein g . In a

C onstitution w ith a) individual rankings o f m ore than tw o alternatives w hich are
com plete and transitive, and the aggregate o f such rankings also being the sam e, b)
requirem ent o f satisfying the P areto criterion, c) the rankin g s being u n affected by
rankings and variations therein concerning alternatives w hich are not included in
the above set, an d d) no one person’s rankings determ in in g the ran k in g s o f all
others. A rrow d em onstrated that no aggregation rule regarding rankings can be
derived (A rrow , 1951), Social choices w ould be a null set. Sen has introduced
restrictions and m odifications o f the C onstitution so as to open up scope for positive
social choices. H e has experim ented w ith induction of cardinality and interpersonal
com parisons o f utilities of d ifferent individuals, quasi-prefercntial orderings w hich
perm it w eak preferences, influences o f one individual on som e others and ethical
norm s like the w eightage for the perspective o f the low est strata o f society, etc.. He
has departed from the P areto criterion by allow ing scope for som e p erso n s to be
worse off. Sen has also underlined the im portance of adding m ore inform ation in the
elucidation o f the conjunctures in the d ifferent social choice situations. A ll this has
widened the dim ension o f social choice theory and linked it w ith ethics and v arious
sorts of political preferences and priorities (Sen, 1970). B ut Sen seem s to slick to
the in d iv idualism -centred basis o f social choice. He is in the B e rg so n -A rro w
tradition.
(2) Social choice theory should boldly confront the problem o f equitable appropriate
distribution arrangem ents to be established in society. Such arrangem ents in w elfare
econom ics should not be taken as given.
(31 Rquitable distribution arrangem ents can be decided by inducting near universally
acceptable ethical norm s or rules. One such ethical norm , favoured by R aw ls and
Sen (as also by A rrow in a private com m unication to Dr. P R . B rahm ananda), is that
decisions on social choices should give the greatest w eight to the low est strata o f
society. H ow ever, there is a difference betw een Raw ls and Sen in the m inim a for all.
The form er em phasises prim aries, especially incom e, and the latter capabilities
(Rawls, 1971).
(4) The goal o f econom ic activity m ay not be the m axim isation o f the utilities o f
individuals, but this presum es certain pre-conditions concern in g the p re-ex isten ce
of capabilities on the part o f all the individuals. The assum ption is that every
individual operates on the frontier o f choice possibilities. T his assum ption is not
valid in developing countries even if the latter are dem ocracies. The cho ice horizon
of all individuals has to be expanded to the frontier. T his can be achieved only by
developm ent o f capabilities o f hum an beings (Sen, 1985). C apability p reced es the

Journal o f Social and Economic Development

Vol. II, No.l, Jan-June 1999
 144 P. R. Brahmananda

process by w hich individuals achieve their w ellbeing potentials. C ap ab ilities are

developed in hum an beings by the intake of/or access to p articu lar categ o ries o f
com m odities. The capability-yielding goods are adequate an d nutritious food and
assurance of their availability in all situations, reasonable m edical and h ea lth c are,
and m inim um elem entary education.
(5) The m arket m echanism by itself w ill not achieve the above req u irem en t o f
capability developm ent for all individuals in develo p in g countries. T his is because
of the absence o f scope for m arket entitlem ents for the above co m m odities for all
persons irrespective of their initial econom ic states, and the possibilities o f failure
of such en titlem en ts under adverse co n d itio n s for the p o o rer se ctio n s o f the
com m unity. G iven the m arket opportunities as supplem ented by state endeavours,
the actual level o f w ellbeing attainm ent is conditioned by the level o f capability
developm ent.
(6) O ne plausible ethical norm , the w eak equity norm , is that o f the obtaining of
m inim um equal capabilities for all human beings and in the light o f the initial inequities
in existing social and other arrangements. The application o f the norm would therefore
im ply som e m easure of inequalities also because o f the plurality o f the spaces in
w hich the concept o f inequality gets bogged dow n (Sen, 1992).
(7) The obtaining of m inim um capabilities for all hum an b eings and assurance of
the entitlem ents in all situations for the sam e should be the goal o f social/State
policy.
(8) F am ines, resulting in m ass deaths, are due not to aggregate deficiency in the
production/supply of foodgrains nor to excess population in relation to production
of foodgrains (Sen, 1981). T hey are due to speculative hoarding, high prices due to
in flatio n, in fo rm atio n and c o m m u n ic a tio n gaps due to ab sen c e o f freedom ,
m aldistributions, and absence o f entitlem ents for food on the part o f the vu'nerable
p o p u latio n . Sen does not attrib u te fam in es to m acro ca u se s lik e inadequate
productions, poor carry-over, inadequate or low im ports, etc.
(9) Fam ines are less probable in dem ocracies than in dictatorships. D em ocracies
are characterised by less gaps in inform ation and com m unication than dictatorships.
They also have m ore scope for expressions o f free o pin io n s and enquiries (Sen.
1990).
(10) Poverty eradication has to take into acco u n t the d istrib u tio n param etcr(s)
below the poverty line. The m easure o f poverty is low er the h ig h er the ratio of the
mean consum ption o f people below the poverty line to the poverty line, and the
less skew ed the distribution o f consum ption below the poverty line (Sen, 1985)
The basic positive contribution o f Sen is to su b stitu te utilitarianism by
capabilitarm nism . Society should have the goal o f d ev elo p in g the capabilities of

Journal o f Social and Economic Development

Vol. If.No.i, Jan-June 1999
 The Welfare Economics o f Amartya Sen 145

individuals. N one should have a level o f capability belo w a norm . C a p ab ility is

treated as an asset endogenous to the individual, com prisin g level o f incom e and
wealth, ability to partake of basic liberties to m ake use o f the facility o f freedom o f
m ovem ent and choice o f occupation, to participate in the process o f en jo y m en t o f
pow ers and prerogative o f offices and positions and resp o n sib ilities open to all
citizens, and to fully utilise the social bases o f self-rcspect will be a function o f the
level o f capabilities o f an individual. M ere constitutional or conventional p ro v isio n
of equal opportunities in society for enjoying the R aw lsian prim aries w ill not by
itself lead to all the individuals m aking use o f the above provisions. If w e treat the
provisions by society as the supply side o f prim aries, the dem and side will d ep en d
on the capabilities o f individuals. Sen w ould probably consider the low level o f
human d evelopm ent in developing econom ies as being caused by the low level o f
capabilities o f the m asses. C apability is not the classical concept o f p ro d u ctiv ity as
understood in econom ics. In the latter w e w ould m ake capability a fu n ctio n o f real
physical capital/land externally with the individual and o f the human capital invested
in him . Sen uses the term capability in a w ider sense to include the in d iv id u al’s
ability to function in society in various capacities, not m erely as a w orker.

Social Choice
Social choice refers to the choices m ade by d ifferen t in d iv id u als in a
com m unity concerning alternative polices, projects and program m es, w hich have
an im pact on the com m unity. A bram B ergson m ade social w elfare a fu n ctio n of the
levels of utilities o f individuals. If the levels o f utilities o f all individuals w ent up by
the adoption o f a p articular policy etc., that policy w as supposed to co n trib u te to
an im provem ent in social w elfare. A lternatively, if som e individuals benefited from
a particular policy, and the other individuals w ere unaffected and did not o b ject to
the above individuals benefiting from the policy, social w elfare w ould still go up.
Note that each individual is taken separately and w e are co ncerned only w ith
whether the levels o f utilities go up and not w ith by how m uch they go up. T h e
Bergson function does not take into account cases w here som e in d iv id u als’ lev els
of utilities go up and those o f som e others g o dow n.
K enneth A rrow , pursuing the sam e individualistic tradition, d ev ised a
hypothetical C onstitution w ith the follow ing conditions:
(1) The Pareto criterion o f im provem ent, w hich is the sam e as Bergson had stipulated
as above.
(2) Each individual has m ore than tw o alternatives. AH the alternatives are p laced
before him. If, for exam ple, he prefers A to B and B to C, he prefers A to C . T hus, he

Journal o f Social and Economic Development

Vol. II, No.l, Jan-June 1999
 146 P. R, Brahmananda

is able to rank all the alternatives and his ranking is transitive. T he sam e is tru e for
the whole com m unity of different individuals. (Individual and Collective rationality.).
(3) No one individual's ranking will dom inate the rankings of everyone else. (N o n ­
dictatorship).
(4) The alternatives w hich are independent of those included in the choice set are
not taken into consideration in the rankings am ong the chosen set by the d ifferen t
individuals. (Independence of irrelevant alternatives).
Take three courses of action — A. B, C — and three individuals — X , Y, Z
or groups, X, Y, Z o f the sam e num ber o f individuals. X prefers A to B and B to C.
Y prefers B to C and C to A. Z prefers C to A and A to B. H ere we have X and Y
preferring B against C; Y and Z preferring C to A and X and Z preferring A to B. No
social choice can be made. An equal m ajority prefers each o f the alternatives.
So long as the alternatives in this set are w holly in the realm of econom ics
and since wc do not know by how m uch by each A or B or C is preferred o v er the
other, if the higher preference of any of the individuals is ex cluded in the ranking,
and if he feels w orse o ff on that score, no choice can be made.
T he problem becom es m ore in tractable if the th ree in d iv id u als have
different values. We cannot rank one value as against another. If the alternatives
involve values or also values along w ith the possibility of eco n o m ic im provem ent,
the person or persons w hose values are ignored in the choice m ay feci w orse off.
H ence, no social choices can be m ade w hen there are differences in values and the
values are interlinked with econom ic benefits in the alternatives.

Efforts to Get Out of the Arrow Groove

Efforts have been m ade to find ways out o f A rrow ’s im possibility of social
choice iheorem by confining the alternatives to w holly econom ic items. Wc assume
eith er that the value system s are hom ogeneous or that the altern ativ es are equally
neutral concerning values.
B y bringing in cardinal utility and interpersonal com parisons of utilities, it
is possible to make social choices. T his was dem onstrated as a general possibility
by Sen, though A rrow him self had accepted this way out. Sen also toyed with a
second way out by show ing that at least in som e choice sets, c a r d i n a l i t y and
interpersonal com parisons can be perm itted. Further, Sen show ed that by introducing
indifference betw een tw o alternatives in the set, one may get over the hurdle. He
also dem onstrated that if one individual influences som e o th ers, though not all. in
som e cases choices can be m ade. The m ore pow erful way by w hich he sought to

Journal o f Social and Economic Development

Vol. II, No.!, Jan-June 1999
 The Welfare Economics ofAmartya Sen 147

show the possibility o f social choice w as by m odifying the C onstitution to allow for
the condition that o f the d ifferent alternatives that w hich benefits the p o o rest or the
w eakest should get higher w eightage. In other w ords, if gro u p Z consists o f a larg er
num ber o f p o o r individuals com pared w ith other groups, if Z prefers C to A and A
to B, its preference w ill get h igher w eight even if there is an equal m ajo rity in term s
o f num bers for A or B or C.

Axiomatic Approach

T he w eightage for the poorest w as indirectly draw n from the co n trib u tio n
o f Raw ls to Justice. Even slightly apart from that, it is possible to form ulate universal
axiom s by m eans o f w hich the social choice im possibility theorem can be m odified.
For exam ple, the rule that in all choices the alternative w hich benefits th e lo w est
groups m ore than the other alternatives are able to can be introduced as an axiom .
But unless we bring in cardinality and interpersonal com parisons, it is not
possible to m ake social choices w hen som e people feel w orse o ff from an econom ic
point of view, w here one alternative is preferred to others.
H ow ever, Sen has been able to show that in m o st other cases w here
differences in values are not concerned, it is possible to find a way out o f the social
choice im possibility theorem by m odifications of the C onstitution. B u t given the
.Arrow conditions o f the C onstitution, it seem s the theorem has its sway. H ow ever,
there is one condition w hich has not been specified in the A rrow C o n stitu tio n .
Suppose in the earlier exam ple there is an alternative w hich is a linear m ix of A, B and
C. very often this m ay be chosen by each of the individuals w hen there is a stalem ate
and it is know n th a t confining the choices to either A or B or C w ould lead to such
a stalemate. A rrow has to rule this out. In m ost com m unities alienation o f any group
is sought to be avoided by all the groups. That is why a lin ear com bination o f the
different alternatives w ould m ost probably be the social ch o ice in a dem ocracy w ith
heterogenous com ponents o f population even w ith different value system . I m ention
this as this way out o f the A rrow stalem ate does not seem to have been co n sid ered
m the literature to the best o f m y judgem ent.
O ne w eakness o f the m ajority basis for social ch o ices is that if th e sam e
majority goes on getting its preferences satisfied, the excluded non-m ajority groups
would get alienated. If the m ajority is a religious, ethnic, language or ideological
group, the process of disintegration o f a com m unity will get strengthened. W icksell
did not accept the 51 % rule for m ajority support. He w anted the base to be enlarged
to over 80 per cent. It is necessary to supplem ent the m ajority basis by a trusteeship
basis in a m uiti-factioned dem ocracy.

Journal o f Social and Economic Development

Vol. HNo . } , Jan-June 1999
 148 P. R. Brahmananda

Capabilitarianism
The axiom atic approach assum es that in regard to certain issues w hich are com m on
to all, there will tend to be scope for universal support for the axiom s. This, of
course, m ay not alw ays be true. In every econom y there are com peting claim s on
resources to be used by the public bodies. T he budget co n strain t in a context of
com peting claim s m akes it difficult to satisfy all parties. Som e w ill necessarily be
excluded. K ant’s categorical im perative can provide a basis for axiom s. K an t’s
approach to ethics may not be accepted by all. T hat individuals seek to adopt the
best values for them selves may be true in an idealist com m unity. Sen has introduced
the w eak equity axiom in cases w here initially som e in d iv id u als or groups of
individuals are in a w orse off state than others in regard to cap ab ilities o f one sort
or the other. Sen w ould argue that those w ho are below a m inim um standard of
capabilities will have to be lifted to the m inim um by discrim in ato ry advantages
socially accorded to them. Equal opportunities for all will not satisfy here. There are
problem s in reducing different types o f capabilities to a com m on yardstick. Second,
there are problem s in com paring one aggregate m easure o f cap ab ility w ith a given
m ix o f different types o f capabilities w ith an o th er aggregate m easure o f a different
m ix of capabilities. How do w e say that one level o f general capability is above that
or below that of another? How do we com pare different in d iv id u als’ capabilities? Is
there a law o f dim inishing m arginal capability or capability m ix?
In the Indian context, the w eak equity axiom m ay lead to reservations.
D oes this axiom hold for all tim e? D oes it hold for every individual w ithin a group?
Is there not a difference betw een rules equity and act equity? C an w e im prove the
cap ab ility of one group w ithout potentially reducing the cap ab ility o f another?
W hat is the philosophical basis for the sam e?
If som e goods are supposed to lead to capability, can n o t the latter itself be
treated as a com m odity? It costs resources to obtain capability. H ow do w e measure
the cost-benefit ratio here? W here do we stop input of reso u rces for obtaining/
im proving capabilities? If capability is desired for its ow n sake and if capability
attainm ent involves scarce resources, w hat is the basis for p riority tow ards that?

Constraints on Capability

John Raw ls justified m inim um incom e for all on the basis o f a h y p o th e tic a l
social contract in a dem ocratic society. But, such m inim um incom e had to b e out o f
a m axim um . In S en’s approach, the m axim um condition is not generally m en tio n ed .
F inancing o f investm ent in capabilities requires taxation. W h at is the basis for such

Journal o f Social and Economic Development

Vol. II, No.!. Jan June 1999
 The Welfare Economics o f Amartya Sen 149

taxation? Do we levy a tax on the m ore capable and give a subsidy to the less
capable? H ow w ill such a tax affect production and grow th?
But R aw ls refers to som e developed societies. In the case o f others, it is a
norm ative concept. W hy only that norm ? Is econom ic ju stice co nfined to that
norm '7 In S e n ’s case also there arc alternative m inim a even in resp ect o f equal
capabilities. S tate subsidy m ig h t reduce parental re sp o n sib ilitie s and d u tie s.
Entitlem ents should be corresponded by duties, according to one school. To w hat
extent is existence o f people w ith capabilities below a norm due to absence o f
parental attention? W hat are the constraints affecting the parents?

The Trusteeship Approach

G andhiji straightw ay accepted the trusteeship concept. B ut this cam e
from his b elief that all resources belong to God, T hose w ho are better o ff had a d uty
tow ards those w ho are w orse o ff and the latter were not exp ected to be je alo u s o f
the form er or covetous o f the same.
In a trusteeship econom y those w ho are m ost w orse off have an entitlem ent
to receive transfers from those w ho are m ost better off. A lternatively, those w ho are
most better o ff have a duty to transfer a portion o f their incom es/w eal th/yields o f
capabilities to those w ho are m ost w orse off.
In a dem ocracy governed by the m ajority opinion /p referen ces, if those
who are m ost w orse off constitute a m inority the la tte r’s interests w ill not be taken
care of by the fo rm e r It is necessary to com plem ent the m ajority basis by the
trusteeship basis.
S uppose we accept the axiom that all those w ho are at the low est rung in
the hierarchy o f capability m easures should be m oved up by m eans o f su bsidies,
etc. Logically, this would imply ultim ately a society of equal m easures o f capabilities
for all. Sim ilarly, if we do not have an absolute m inim um incom e, etc., line and ju s t
accept the axiom that all those at the low est rungs will be continuously m oved up,
in the limit, w e end up w ith equal incom es for all. A ny equity axiom w ithout a
specification o f a m inim um will land in an equality axiom . If wc do not constrain any
such axiom by a requirem ent that aggregate production, productivity etc., should
not be reduced, actually or potentially by the application o f the axiom , society may
become poorer and poorer, and efficiency m ight be sacrificed at the co st o f equity.
The maximin of Raw ls has accepted this constraint, but the w eak equity axiom does
not seem to have im posed any constraint in its application.
Since capability exists in several spaces and d ifferen t in d iv id u als have
capability gaps in different spheres, we have to specify m inim um equal capability in

Journal o f Social and Economic Development

Vol. II, No.l, Jan-June 1999
 150 P R. Brahmananda

every space and for every individual. H ow do w e define capability in the abstract so
that it encom passes all the spaces and o f all individuals? We m ay not define
capability w ith reference to specific item s.

Capability and Utility

Sen seem s to treat capability on a par w ith utility. C apabilitarianism would
em phasise creation o f potential abilities in a num ber o f relevant econom ic respects.
Since it refers to the ability to function in several capacities, it requires m easu res to
w ork out potential abilities o f different individuals in different spheres. S ince it is
not put up as a goal desired by individuals in their econom ic fu n ctio n in g and is
em ph asised as a norm ative requirem ent specified by the philosopher, it m ay not
explain actual situations. U tilitarianism is em phasised with respect to effects and
consequences. W hat is the counterpart of this to capabilitarianism ? Do we consider
it as a prerequisite for successful utilitarianism ? In such a case, it cannot be defended
on the ground o f pure ethical considerations as desirable for its ow n sake irrespective
o f consequences. If it is treated on a par w ith utilitarianism , then consequences
alone m atter and not intentions or m otives. Take a concrete case, ed ucation m ay be
defended as desirable for its ow n sake. It m ay be also defended on the score o f its
p robable beneficial effects on society, in w hich case it com es u n d er the utilitarian
um brella. Here, w e have to consider alternatives to outlays on education and uphold
that alternative w hich has the m ost beneficial consequences. T here will be com peting
dem ands on resources for provision o f education for all, h ealth care for all, nutritious
food for all, etc. N ote one very im portant im plication o f this procedure. W e cannot
bring capability then under the axiom atic approach unless w e specify the general
rule that resources should be used only for those purposes w hich have probably
net beneficial effects and in a hierarchical m anner. T he same p rocedure as under the
cost-benefit approach becom es relevant and capabilitarianism becom es no different
from utilitarianism.
My jud g em en t, and I stand to be corrected, is that Sen has not yet driven
h im se lf to the c o m er on w hether he places the capability d o ctrin e on the K a n t ia n
ethical basis or on the utilitarian basis. Further, he has not d issected the concept ot
cap ab ility to eschew the utilitarian com ponents in it. He has also not e x a m i n e d
social constraints he w ould like to place on public expenditures on different types
o f capability. Probably, m ore w ork needs to be done to estab lish it on a secure
foundation suchw ise as to d ifferentiate it from the utility concept. I f w e define
capability as leading, though indirectly, to utilities as in th e case of c o n s u m p t i o n

Journal o f Social and Economic Development

Vol. II, No.I, Jm -June 1999
 The Welfare Economics o f Amartya Sen 151

goods, capital goods, profit, interest, etc., w e have to ask w hat is the net ad v an ce to
theory by that concept?

Entitlements
Sen has advanced the view that the attainm ent o f m inim um cap ab ilities
w ould require individuals to be given entitlem ents for the sam e. E n titlem en ts are
legal rights for citizens to claim from the State access to the capability y ield in g
goods and services. E ntitlem ents do not im pose duties and responsibilities. C learly,
the entitlement approach leads to com m itm ent o f society’s resources for the provision
o f supplies o f the above sorts of goods and services. N aturally, there is a b u d g et
constraint to the governm ent. If this constraint is to be lifted, new and additional
taxes have to be im posed. A s noted ea rlier, p ro g re ssiv e ta x atio n w ill en tail
proportionately m ore burdens on the m ore capable than on the less cap ab le. N ow ,
capability here has to be appraised in term s o f incom e or w ealth. U nless the axiom
is accepted by the better to do, there is resistance to m ore o f taxation. T his is ap art
from the leakage in the delivery o f the capability yielding com m odities. Sen has not
addressed these issues. T h ere are also ethical issues co ncern in g the d em arcatio n
of the private spaces of parents vis-&-vis that o f the State.
Since entitlem ents do not im ply responsibilities o r duties, the co st-b en efit
approach cannot be applied here. Further, how do we com pel citizens to use the
entitlem ents? Should we force children to go to school? T his m eans the exercise o f
the S tate’s coercive pow er on the parents or relatives. D o w e use incentives for the
same? In that case en titlem ents are not enough. Probably entitlem ents w ill be used
mostly by those w ho can oth erw ise afford the facilities. E n titlem en ts w ithout
coercion may not adequately fulfil them selves in societies as in Bihar, UP, etc.
Probably, duties will have to be im posed along w ith entitlem ents. In that case,
unlimited freedom s w ill have to be restricted in the social interests. T h at w ould
perhaps be G an d h iji’s view. In an educated and literate com m unity, en titlem en ts
might suffice. But, in m any cases, w ithout duties, the entitlem ent approach m ay not
satisfy or achieve m uch. I h ave not referred to the transaction costs in citizens
getting their entitlem ents enforced. T hese could be significan t for in d iv id u als in a
poor country.

The Bengal Famine

In the case o f the B engal fam ine, it is difficult to accep t the p o sitio n that
liri!y inflation in the absence o f entitlem ents caused the large deaths. In flatio n can

Journal o f Social and Economic Development

Vol. II, No. I J a n J u n e 1999
 152 P. R, Brahmananda

occur w ithout fam ines. A nd inflation in the absence o f index-linking o f all w ages
and incom es in a context o f aggregate shortages can cause fam ines. In the case o f
Bengal. S en ’s data indicates that relative to rice, the prices o f fish and w heat were
falling. Clearly, fam ine deaths w ould have been less had the su bstitution effect
worked, though the incom e effect had m eant loss of real purchasing power. Rigidities
in food habits, w hich have disappeared in W est B engal of today, m ay have brought
m ore fam ine deaths. From a historical angle, for m ore than four decades, there was
no fam ine in Bengal. From a statistical probability angle, it w ould have been difficult
to m ake a case for the holding o f large stocks by the G ov ern m en t o f Bengal for the
earlier years. The Bengal governm ent refused point blank to co n sid er rationing for
the w hole o f Bengal in their evidence before the F am ine E nquiry C om m ission.
Probably, for the Bengal fam ine case, it w ould be difficult to contend that aggregate
deficiency o f food w as one of the leading causes o f the fam ine deaths.
G iven any institutional constraints w hich may com e from history or the
contem porary situation, it is only through im provem ents in aggregate supplies that
large deaths due to sudden fam ines m ay be prevented. A n ticip ated fam ines can be
countered through prior thought out m easures.

Inequality and Poverty

S en ’s im portant contribution on the relation betw een inequality and poverty

is his em phasis on the m easure of inequality of persons below th e poverty line. The
poverty gap m easure has to be m agnified by adjustm ent for the m easure o f inequality
below the poverty line. It is interesting that detailed tim e-series for w hatever they
are w orth, do not alw ays indicate a positive relation betw een th e head-count ratio
and the poverty gap ratio. N or is there alw ays a positive relation betw een the head­
count ratio and the Gini coefficient. S e n ’s contention that poverty is, by and large,
due to a failure o f society on the capability front also may not alw ays be em pirically
supported, if we define capability w ith reference to prim ary ed ucation, health care
and nutritious food.
Conclusion
There is no doubt that Sen has broadened the agenda o f w elfare economics.
He has caused a w elcom e shift in w elfare econom ics in favour o f the problem s of
the poorer and less developed countries. He has probably raised m ore f u n d a m e n t a l
questio n s in the latter area than m ost other econom ists have. T hese questions do
d istu rb the conscience of econom ists all over the w orld. A n sw ers and practical

Journal o f Social and Economic Development

Vol. II, No. I, Jan J u n e ! 999
 The Welfare Economics o f Amartya Sen 153

solutions belong to the realm of political econom y and not to that o f eco n o m ic
theory as such.

References
Arrow. K.J. (1951). S o cia l Choice a n d In d ivid u a l Values. II e d itio n , N ew
York: Wiley, 1963.
Rawls, J. (1971), A Theory o f Justice. Oxford: Clarendon Press.
Sen, A.K. (1970). C ollective Choice and Social W elfare. L ondon: O liv e r and
Boyd.
— (1981). P o ve rty and Fam ines: An E ssay on E n titlem en t a n d D e p riv a tio n .
Oxford; Clarendon Press.
— (1985). Commodities and Capabilities (Hennipman Lecture of 1982). Amsterdam;
Elsevier Science.
— (1990). Socialism , Markets and Democracy. Indian Economic Jo u rn a l, A pril-
June.
— (1992). Inequalities Reexamined. Oxford: Clarendon Press.

Journal o f Social and Economic Development

Vol. 11, Nod, Jan-June 1999