OXFORD READINGS IN FEMINISM
Edited by Anne Phillips
VIJ)EM0CRAC\/ 2^ RfcpKESEkfrAipOM'-OR,  WHV iT VAATTeA WHO 0^ ieep^£S~eNlTA-TlV£>
Oxford • New York
OXFORD UNIVERSITY PRESS
1998
1
Democracy and Representation: Or, Why Should it Matter Who our Representatives Are?
Anne Phillips
Though the overall statistics on women in politics continue to tell their dreary tale of under-representation, this under-representation is now widely regarded as a problem, and a significant number of political parties have adopted measures to raise the proportion of women elected. That the issue is even discussed marks a significant change. Even more remarkable is that growing support for a variety of enabling devices (day-schools, for example, to encourage potential women candidates) now combines with some minority backing for measures that guarantee parity between women and men. Parties in the Nordic countries took the lead in this, introducing gender quotas for the selection of parliamentary candidates from the mid-1970s onwards, but a quick survey across Europe throws up a number of parallel developments. Positive action to increase the proportion of women elected is now on the political agenda. It has become one of the issues on which politicians disagree.
In some ways, indeed, this is an area where those engaged in the practice of politics have edged ahead of those engaged in its theory. Gatherings of party politicians are significantly more likely to admit the problem of women's under-representation than gatherings of political scientists, for while the former remain deeply divided over the particular measures they will support, most can manage at least a lukewarm expression of'regret' that so few women are elected. The pressures of party competition weigh heavily on their shoulders. In an era of increased voter volatility, they cannot afford to disparage issues that competitors might turn to electoral advantage. Hence the cumulative effect noted in Norwegian politics, where the Socialist Left Party first adopted gender quotas in the 1970s; this was followed in the 1980s by similar initiatives from the Labour and Centre
First published in Schweizerisches Jahrbuch für Politische Wissenschaften (1994), 63-76. Reprinted by permission of Editions Paul Haupt Berne. An expanded version of this argument is in The Politics of Presence (Oxford University Press, 1995).
DEMOCRACY AND K^PRESENTATION
Parties; and was accompanied by substantial increases in the number of women selected by the Conservative Party as well (Skjeie 1991). Hence the impact of the German Green Party, which decided to alternate women and men on its list for the 1986 election; the threat of this small—but at the time, rapidly growing—party contributed to the Christian Democrats' adoption of a voluntary quota, and the Social Democrat's conversion to a formal one (Chapman 1993: ch. 9). Hence the otherwise surprising consensus that has emerged among Britain's major political parties—at central office level if not yet in local constituencies—in favour of selecting a higher proportion of women candidates (Lovenduski and Norris 1989). None of this would have happened without vigorous campaigning inside the political parties, but the campaigns have proved particularly effective where parties were already worried about their electoral appeal.
This pragmatically driven conversion contrasts with a more tough-minded resistance inside the political science community, where arguments range from a supposed lack of evidence that sex affects policy decisions, to a distaste with what is implied in saying that it should. Women's under-representation in politics is in one sense just empirical fact; they are not present in elected assemblies in the same proportions as they are present in the electorate. But the characteristics of those elected may diverge in any number of ways from the characteristics of those who elect them, and this is not always seen as a matter of democratic consequence. In a much cited article on representation, A. Phillips Griffiths (1960: 190) argued that some divergences are regarded as positively beneficial. We do not normally consider the interests of lunatics as best represented by people who are mad, and 'while we might well wish to complain that there are not enough representative members of the working class among Parliamentary representatives, we would not want to complain that the large class of stupid or maleficent people have too few representatives in Parliament: quite the contrary'. Feminists may find the implied parallels unconvincing, especially when we recall the many decades in which women were classified with children and the insane as ineligible for the right to a vote; but the general point remains. Establishing an empirical under-representation of certain categories of people does not in itself add up to a normative case for their equal or proportionate presence. It may alert us to overt forms of discrimination that are keeping certain people out, but does not yet provide the basis for radical change.
The contemporary version of Phillips Griffiths' argument takes
224
225
ANNV. I'HIUJI'S
the form of the notorious 'slippery slope': if measures are proposed for achieving a fair 'representation' of the proportion of women in the electorate, why not also of homosexuals, of pensioners, of the unemployed, of people with blue eyes and red hair? Though usually raised with deliberately facetious intent, such questions combine with more serious work on representation which has tended to dismiss ideals of 'descriptive' or 'mirror' representation as a nostalgic yearning for direct democracy. In her influential work on The Concept of Representation, Hanna Pitkin (1967; 86) suggests that the metaphors of descriptive representation are most commonly found among those who regard representative democracy as a poor second-best, and who therefore look to more 'accurate' or pictorial representation of the electorate as a way of approximating the old citizen assemblies. Yet representatives, she argues, are supposed to act'—what would be the point of a system of representation that involved no responsibility for delivering policy results?—and too much emphasis on who is present may divert us from the more urgent questions of what the representatives actually do, 'Think of the legislature as a pictorial representation or a representative sample of the nation, and you will almost certainly concentrate on its composition rather than its activities' (1967: 226). In Pitkin's preferred version, it is the activities rather than the characteristics that matter, and what happens after the action rather than before it that counts. Representing 'means acting in the interests of the represented, in a manner responsive to them' (1967: 209). Fair representation cannot be guaranteed in advance; it is achieved in more continuous process, which depends on a (somewhat unspecified) level of responsiveness to the electorate. The representatives may and almost certainly will differ from those they act for, not only in their social and sexual characteristics, but also in their understanding of where the 'true' interests of their constituents lie. What renders this representative is the requirement for responsiveness. 'There need not be a constant activity of responding, but there must be a constant condition of responsiveness, of potential readiness to respond' (1967: 233).
Radicals may challenge this resolution as allowing too much independence of judgement and action to the representatives, but the direction their criticisms take also lends little support to arguments for gender parity. The most radical among them will scorn what they see as a reformist preoccupation with the composition of political elites—and they may express some dismay that a once obsessively democratic women's movement could retreat to such
DEMOCRACY AND REPHESENTATION
limited ambitions. Others will give more serious consideration to reforms that increase the representative nature of existing national assemblies, but they will prefer mechanisms of accountability that minimize the significance of the individuals elected. The shift from direct to representative democracy has shifted the emphasis from who the politicians are to what (policies, preferences, ideas) they represent, and in doing so, has made accountability to the electorate the pre-eminent concern. We may no longer have much hope of sharing in the activities of government, but we can at least demand that our politicians do what they promised to do. The quality of representation is then thought to depend on tighter mechanisms of accountability that bind politicians more closely to the opinions they profess to represent. Where such processes are successful, they reduce the discretion and autonomy of individual representatives; in the process, they seem to minimize the importance of whether these individuals are women or men.
Consider, in this context, the guidelines that were introduced by the US Democrats in the early 1970s, to make their National Convention (which carries the crucial responsibility of deciding on the presidential candidate) more representative of the party rank and file. Dismay at the seemingly undemocratic nature of the 1968 Convention prompted the formation of a Commission on Party Structure and Delegate Selection, which recommended more extensive participation by party members in the selection of delegates, as well as quota guidelines to increase the proportion of delegates who were female, black, and young. As a result of this, the composition of the 1972 Convention was markedly more 'descriptive' of Party members than previous ones had been: 40 per cent of the delegates were women, 15 per cent were black, and 21 per cent were aged between 18 and 30. But the reforms pointed in potentially contradictory directions, for they simultaneously sought to increase rank and file participation in the selection of delegates, to bind delegates more tightly to the preferences of this rank and file, and to ensure a more descriptive representation according to age, gender, and race. As Austin Ranney (1982: 196)—one of the members of the Commission—later noted, the success of the first two initiatives undermined the importance of the third. By 1980, the overwhelming majority of delegates were being chosen in party primaries which bound them to cast their votes for one particular candidate; they became in consequence mere ciphers, who were there to register preferences already expressed. 'If that is the case,' Ranney argues, 'then it really doesn't matter very much who the delegates are,' The
226
227
ANNE PHILLIPS
more radical the emphasis on accountability, the less significance attaches to who does the work of representation.
Those engaged in campaigns for gender quotas have worked with some success on the electoral sensitivities of party politicians, but have made less headway among the tough-minded theorists of representation. My concern here is to address the latter, and to create maximum difficulties for myself I will focus on the stronger claim of gender parity, rather than the more modest claim for some more women elected.1 This reflects what may be a naive confidence on my part: that no one who seriously considers the matter could regard the current balance between the sexes as a fair process of representation. At the lowest points of women's under-representation (it was only in 1987 for example, that the British House of Commons lifted itself above the 5 per cent mark), one need only reverse the position of the sexes to demonstrate the democratic deficit. What would men think of a system of political representation in which they were outnumbered nineteen to one? At such gross levels of gender imbalance, rhetorical devices are all that we need—one would have to be a pretty determined patriarch to defend this as an appropriate state of affairs. But recent initiatives have raised the stakes considerably higher, insisting on positive action as a condition for effective change, and aiming at fifty/fifty parity, or a 40 per cent minimum for either sex. What are the arguments for this more radical position, and how do they engage with current conventions of accountability and representation?
Arguments for raising the proportion of women elected fall broadly into four groups. There are those that dwell on the role model successful women politicians offer; those that appeal to principles of justice between the sexes; those that identify particular interests of women that would be otherwise overlooked; and those that point towards a revitalized democracy that bridges the gap between representation and participation. The least interesting of these, from my point of view, is the role model. When more women candidates are elected, their example is said to raise women's self-esteem, encourage others to follow in their footsteps, and dislodge deep-rooted assumptions on what is appropriate to women and men. I leave this to one side, for I see it as an argument that has no particular purchase on politics per se. Positive role models are certainly beneficial, but I want to address arguments that engage more directly with issues of democracy and representation.
One final preamble. Though I deal here only with general issues of justification, there is a second order question, which is how legit-
228
DEMOCRACY AND REPRESENTATION
imate objectives can be best achieved. The emphasis on quota mechanisms and other such guarantees has aroused strong resistance even among those who claim to share the ultimate goal of women's equality in politics, and while some of this can be discounted as intellectual or political dishonesty, much of it relates to pragmatic judgements of what is possible in any particular context. The potential backlash against women is one consideration here, as are the difficulties some political parties claim to experience in finding enough women candidates. Some of the resistance depends on more general arguments against positive action; some of it reflects still unresolved tensions between gender and class; some relates to a familiar problem in political argument, which is that mechanisms proposed for achieving one desired goal can conflict with other desirable ends. Considerations of space prevent me dealing with this second order question, and I will merely note that there are pragmatic judgements to be made, which do not flow simply from the conclusions on general objectives. But if gender parity can be shown to matter, and existing structures can be shown to discourage it, this constitutes a case for positive action.
I. THE CASE FOR GENDER PARITY: THE JUSTICE ARGUMENT
One of the most powerful arguments for gender parity is simply in terms of justice: that it is patently and grotesquely unfair for men to monopolize representation. If there were no obstacles operating to keep certain groups of people out of political life, then we would expect positions of political influence to be randomly distributed between both sexes and across all the ethnic groups that make up the society. There might be some minor and innocent deviations, but any more distorted distribution of political office is evidence of intentional or structural discrimination (Phillips 1991). In such contexts (that is, most contexts!) women are being denied rights and opportunities that are currently available to men. There is a prima facie case for action.
There are three things to be said about this argument. One is that it relies on a strong position on the current sexual division of labour as inequitable and 'unnatural'. Consider the parallel under-representation of the very young and very old in politics. Most people will accept this as part of a normal and natural life-cycle, in
229
ANN I; I'HI 1.1,11'S
which the young have no time for conventional politics, and the old have already contributed their share; and since each in principle has a chance in the middle years of life, this under-representation does not strike us as particularly unfair. The consequent 'exclusion' of certain views or experiences may be said to pose a problem. But however much people worry about this, they rarely argue for proportionate representation for the over-70s and the under-25s.2 The situation of women looks more obviously unfair, in that women will be under-rep resented throughout their entire lives, but anyone wedded to the current division of labour can treat it as a parallel case. A woman's life-cycle typically includes a lengthy period of caring for children, and another lengthy period of caring for parents as they grow old. It is hardly surprising, then, that fewer women come forward as candidates, or that so few women are elected. Here, too, there may be an under-representation of particular experiences and concerns, but since this arises quite 'naturally' from particular life-cycles it is not at odds with equality or justice.
I do not find the parallel convincing, but my reasons lie in a feminist analysis of the sexual division of labour as 'unnatural' and unjust. The general argument from equal rights or opportunities only translates into a specific case for gender parity in politics when it is combined with some such analysis; failing this, it engages merely with the more overt forms of discrimination that exclude women from political office, and cannot deliver any stronger conclusion. Justice requires us to eliminate discrimination (this is already implied in the notion of justice), but the argument for women's equal representation in politics depends on that further ingredient which establishes structural discrimination. Feminists will have no difficulty adding this. This first point then helps clarify what is involved in moving from a description of women's under-representation to an analysis of its injustice.
The second and third points are more intrinsically problematic, and relate to the status of representation as a political act. If we treat the under-representation of women in politics as akin to their under-representation in management or the professions, we seem to treat being a politician as on a continuum with all those other careers that should be opened up equally to women. In each case, there is disturbing evidence of sexual inequality; in each case, there should be positive action for change. The argument appeals to our sense of justice, but it does so at the expense of an equally strong feeling that being a politician is not just another kind of job. 'Career politician' is still—and rightly—a term of abuse; however accurately
DEMOCRACY AND REPRESENTATION
it may describe people's activities in politics, it does not capture our political ideals. If political office has been reduced to yet another favourable and privileged position, then there is a clear argument from justice for making such office equally available to women. Most democrats, however, will want to resist pressures to regard political office in this way. So while men have no 'right' to monopolize political office, there is something rather unsatisfying in basing women's claim to political equality on an equal right to an interesting job.
An alternative and more promising formulation considers the under-representation of women in elected assemblies as analogous to their under-representation in the membership of political parties or the attendance at political meetings, and thus treats the equal right to be an elected representative as part of an equal right to political participation. This provides a more theoretically satisfying foundation, for equality in participation is one of the criteria by which democracies are judged, and the systematic under-participation of particular social groups is normally regarded as a political problem (Verba, Nie, and Kim 1978; Parry, Moyser, and Day 1992). This is not to say that everyone must be equally enthralled by the political process: the interest in politics is unevenly distributed, as is the interest in sport or in jazz. But when the distribution coincides too neatly with divisions by class or gender or ethnicity, political participation is by definition unequal and political influence as a consequence skewed. The principle of a rough equality between various social groups is already implicit in our idea of participation, and too marked a deviation from this is already regarded as a political failing. Once gender is admitted as an additional and relevant imbalance, it is easy enough to argue for equal participation between women and men.
As applied to representation, however, the argument seems to assert what has still to be established: that representation is just another aspect of participation, to be judged by the same criteria. Yet many theorists of democracy proceed from just the opposite direction, and they have based much of their critique of direct or participatory democracy on precisely what differentiates representation from participation. Participation implies activity, and yet activity is always a minority affair. By setting the requirements for participation impossibly high, theorists of participatory democracy are said to promote a politics that becomes 'unrepresentative' and unequal, for while most citizens can manage an occasional foray into the polling booth, few are willing or able to take on more
230
231
ANNH PHILLIPS
continuous engagement, and the power then slips into the hands of those who most iove politics. Representative democracy claims to solve this conundrum by removing the requirement for physical presence. As long as there is a minimal level of equality in the act of voting, then the representation can be said to be equal; we do not have to commit ourselves additionally to the hard labour of the political life.
Equality of presence—a rough approximation to the social groups that make up the society—is already implicit in the notion of participation. But it is not so obviously implicit in the notion of representation, which was, if anything, dreamt up to get round this bothersome condition. The two are, of course, related, for a society that provided genuinely equal access to participation in meetings and pressure groups and parties would almost certainly produce the same kind of equality among the people elected. In principle, however, they are separate, for in distancing itself from participating democracy, representative democracy has distanced itself from physical presence as the measure of political equality. Representative democracy claims, for example, to represent the competing interests of capital and labour by giving each of us an equal right to vote, and this is said to encourage a variety of parties to emerge that will speak to our different concerns. But representative democracy makes no claims about achieving a proportionate representation of working class people inside the legislative assemblies: workers, should be equally represented, but not necessarily by workers themselves. So while we can readily appeal to existing understandings of democracy as the basis for women's equal participation, the case for gender parity among elected representatives moves onto more unchartered ground.
What we can perhaps do is turn the argument around, and ask by what 'natural' superiority of talent or experience men could claim a right to dominate assemblies? The burden of proof then shifts to the men, who would have to establish either some genetic distinction which makes them better at understanding problems and taking decisions, or some more socially derived advantage which enhances their political skills. Neither of these looks particularly persuasive: the first has never been successfully established; and the second is no justification if it depends on structures of discrimination. There is no argument from justice that can defend the current state of affairs; and in this more negative sense, there is an argument from justice for parity between women and men. But there is still a troubling sense in which the argument overlooks what is peculiar to represen-
232
DEMOCRACY AND REPRESENTATION
tation as a political act. When democracy has become largely a matter of representing particular policies or programmes or ideas, this leaves a question-mark over why the sex of the representatives should matter.
II. THE CASE FOR GENDER PARITY: WOMEN'S INTERESTS
The second way of arguing for gender parity is in terms of the interests that would be otherwise discounted: this is an argument from political realism. In the heterogeneous societies contained by the modern nation state, there is no transparently obvious 'public interest', but rather a multiplicity of different and potentially conflicting interests which must be acknowledged and held in check. Our political representatives are only human, and as such they cannot pretend to any greater generosity of spirit than those who elected them to office. There may be altruists among them, but it would be unwise to rely on this in framing our constitutional arrangements. Failing Plato's solution to the intrusion of private interest (a class of Guardians with no property or family of their own) we must look to other ways of limiting tyrannical tendencies, and most of these will involve giving all interests their legitimate voice.
This, in essence, was James Mill's case for representative government and an extended franchise, though he notoriously combined this with the argument that women could 'be struck off without inconvenience' from the list of potential claimants, because they had no interests not already included in those of their fathers or husbands. (He also thought we could strike off 'young' men under forty years of age.) Part of the argument for increasing women's political representation looks like a feminist rewrite and extension of this. Women occupy a distinct position within society: they are typically concentrated, for example, in lower-paid jobs; and they carry the primary responsibility for the unpaid work of caring for others. There are particular needs, interests, and concerns that arise from women's experience, and these will be inadequately addressed in a politics that is dominated by men. Equal rights to a vote have not proved strong enough to deal with this problem; there must also be equality among those elected to office.
At an intuitive level, this is hard to fault. It takes what is a widely accepted element in our understanding of democracy and applies it
233
ANN!-; PHILLIPS
to women's situation. Looked at more closely, however, the argument from women's interests or women's concerns seems to rest on three conditions: that women have a distinct and separate interest as women; that this interest cannot be adequately represented by men; and that the election of women ensures its representation. As critics of gender quotas will be quick to point out, each condition is vulnerable to attack. The notion that women have at least some interests distinct from and even in conflict with mens is relatively straightforward (we can all think of appropriate examples), but this falls a long way short of establishing a set of interests shared by all women. If interests are understood in terms of what women express as their priorities and goals, there is considerable disagreement among women, and while attitude surveys frequently expose a 'gender gap' between women and men, the more striking development over recent decades has been the convergence in the voting behaviour of women and men. There may be more mileage in notions of a distinct woman's interest if this is understood in terms of some underlying but as yet unnoticed 'reality', but this edges uncomfortably close to notions of 'false consciousness', which most feminists would prefer to avoid. Indeed the presumption of a clearly demarcated 'woman's interest' which holds true for all women in all classes and all countries has been one of the casualties of recent feminist critique, and the exposure of multiple differences between women has undermined more global understandings of women's interests and concerns (see, for example, Mohanty 1992). If there is no clearly agreed and recognized 'women's interest', does it really matter if the representatives are predominantly men?
Definitive as this might seem, it does not seriously undermine the claim to gender parity; if anything, it can be said to strengthen it. Consider, in this context, Edmund Burke's rather odd understanding of interests as reflecting 'an objective, impersonal, unattached reality', which can then be represented by any sufficiently competent and honest individual (Pitkin 1967: 168). Odd as this is, it conveys a partial truth. The more fixed the interests, or the more definite and easily defined, the less significance attaches to who does the work of representation. So if women's interests were transparently obvious to any intelligent observer, there might be no particular case— beyond the perennial one of trust—for insisting on representatives who also happen to be women. We might feel that men will be less diligent in pressing women's interests or concerns, but if we all know what these are, it will be correspondingly easy to tell whether they are being adequately pursued. If, however, the interests are varied,
DEMOCRACY AND RRPRESliNTATlON
unstable, perhaps still in the process of formation, it will be far more difficult to separate out what is to be represented from who is to do the representation. The greater problems arise, that is, where interests are not so precisely delineated, where the political agenda has been constructed without reference to certain areas of concern, or where much fresh thinking is necessary to work out the appropriate policies. To this extent, the very difficulties in defining what are in women's interests strengthen the case for more women as representatives.
The more decisive problem lies in the third condition. Does the election of more women then ensure their representation? Again, at an intuitive level, an increase in the number of women elected seems likely to change both the practices and the priorities of politics, increasing the attention given to matters of childcare, for example, or ensuring that women's poor position in the labour market is more vigorously addressed. This intuition is already partially confirmed by the experience of those countries which have changed the gender composition of their elected assemblies. But what does this mean in terms of political representation? Elections are typically organized by geographical constituencies, which sometimes coincide with concentrations of particular ethnic or religious groups, or concentrations of certain social classes, but which never coincide with concentrations of women or men. Elections typically take place through the medium of political parties, each of which produces candidates who are said to represent that party's policies and programmes and goals. In what sense can we say that the women elected through this process carry an additional responsibility to represent women? In the absence of mechanisms to establish accountability, the equation of more women with more adequate representation of women's interests looks suspiciously undemocratic. How do the women elected know what the women who elected them want? By what right do they claim responsibility to represent women's concerns?
Though this is rarely stated in the literature, the argument from women's interests implies that representatives will have considerable autonomy: that they do have currently, and by implication, that this ought to continue. Women's exclusion from politics is said to matter precisely because politicians do not abide by pre-agreed policies and goals. As any observer of the political process knows, policy decisions are not settled in advance by party programmes, for new problems and issues emerge alongside unanticipated constraints, and in the subsequent weighing of interpretations and priorities, it
234
235
ANNL;. PHILLIPS
DEMOCRACY AND R Kl> RF.SENTATI ON
matters immensely who the representatives are. Feminists have much experience of this, gained through painful years of watching hard won commitments to sexual equality drop off the final agenda. When there is a significant under-representation of women at the point of final decision, this can and does have serious consequences, and it is partly in reflection of this that feminists have shifted their attention from the details of policy commitments to the composition of the decision-making group. Political experience tells us that all male or mostly male assemblies will be poor judges of women's interests and priorities and concerns, and that trying to shore up this judgement by pre-agreed programmes has only limited effect. There is a strong dose of political realism here. Representatives do have considerable autonomy, which is why it matters who those representatives are.
It is worth dwelling on this point, for it highlights a divergence between current feminist preoccupations and what has long been the main thrust in radical democracy. Radical democrats distrust the wayward autonomy of politicians and the way they concentrate power around them, and they typically work to combat these tendencies by measures that will bind politicians more tightly to their promises, and disperse over-centralized power. Feminists have usually joined forces in support of the second objective: feminism is widely associated with bringing politics closer to home; and women are often intensely involved in local and community affairs. But when feminists insist that the sex of the representatives matters, they are expressing a deeper ambivalence towards the first objective. The politics of binding mandates, for example, turns the representatives into glorified messengers: it puts all the emphasis onto the content of the messages, and makes it irrelevant who the messengers are. In contesting the sex of the representatives, feminists are querying this version of democratic accountability.
The final point about the argument from interests is that it may not of itself justify equal or proportionate presence. In a recent discussion of demands for group representation in Canada, Will Kymlicka (1993) makes a useful distinction between arguments for equal or proportionate presence (where the number of women or aboriginal Indians or francophone Canadians in any legislative assembly would correspond to their proportion in the citizenry as a whole), and the case for a threshold presence (where the numbers would reach the requisite level that ensured each group's concerns were adequately addressed). When the group in question is a numerically small minority, the threshold might prove larger than
their proportion in the population as a whole; when the group composes half the population, the threshold might be considerably lower. On this basis, there might be an argument for greater than proportionate representation of Indians, for example, but less than proportionate representation of women: not that women would be formally restricted to 25 or 30 per cent of the seats, but that they might not require any more than this in order to change the political agenda. It is the argument from justice that most readily translates into strict notions of equality; the argument from women's interests need not deliver such strong results.
III. THE CASE FOR GENDER PARITY:
TOWARDS A REVITALISED DEMOCRACY
The third argument is less developed, and I offer it here as a way of dealing with some of the problems I identify above. The argument from justice works well enough on the limited ground that treats being a politician like any other kind of job, or on the negative ground that denies any just basis for a male monopoly. The argument from women's interests works well enough as a case for a threshold, but not necessarily equal, presence, but is best understood in terms of a realistic assessment of how rarely politicians abide by their pre-agreed programmes. These are powerful arguments, but they are not, on the whole, the kinds of arguments that feminists most admire: they are too much grounded in an impoverished experience of democracy to bear the full weight of feminist ambition. And they leave unresolved that recurrent radical concern about controlling wayward politicians. Apart from the argument that women should get an equal chance at a political career {which is a fair enough argument, but not intrinsically about democracy), we can only believe that the sex of the representatives matters if we think it will change what the representatives do. In saying this, we seem to be undermining accountability through party programmes. We are saying we expect our representatives to do more—or other— than they promised in the election campaign.
There is often an expectation, for example, that women politicians will operate on a cross-party basis, forging alliances to press for improvements in childcare provision or changes in the abortion laws. In her study of Norwegian representatives, HegeSkjeie (1991) records a number of such initiatives, but she notes that it is the
236
237
ANNE PHILLIPS
priorities of their party that finally dictate the way women politicians vote. If we are either surprised or disappointed by this, it must be because we see an increase in the number of women politicians as challenging the dominance of the party system or the tradition of voting along party lines. Those who feel that the tighter controls of party discipline have discouraged serious discussion and debate may be happy enough with this conclusion. But in the absence of alternative mechanisms of consultation or accountability, it does read like a recipe for letting representatives do what they choose to do.
What makes sense of this, I believe, is an additional presumption that is implicit in most feminist arguments, a conviction that changing the composition of existing elected assemblies is only part of a wider project of increasing and enhancing democracy. When the argument for gender parity is taken out of this context, it has to rely more heavily on arguments from political realism, and while these are powerful enough arguments in themselves, they fall short on some key concerns. Put back into its context, the argument often reveals a more ambitious programme of dispersing power through a wider range of decision-making assemblies, and changing the balance between participation and representation.
We might think here of the further initiatives that are so typical of women in politics: the use of the open forum, for example, as a way of consulting women in a local community; the report back to women's sections or women's conferences; or just the extraordinary energy so many women politicians devote to what they see as their responsibilities for representing women. Even among those most committed to party politics (and many women deliberately stay outside, in the more amorphous politics of women's movement groups and campaigns), the political party is frequently viewed as an inadequate vehicle for representation. In 1980s Britain, for example, there was a flowering of women's committees within the framework of local government (usually associated with more left-wing Labour councils), and these made extensive use of co-option or the open forum as a way of consulting women outside the political parties. Now you could think of this as a short term compensation for women's current under-representation among elected councillors, but there is little to support this view. More commonly, those associated with the development of women's committees saw the additional mechanisms of consultation and participation as always and everywhere desirable—even under some future scenario where women might hold 50 per cent of council seats. The women involved were querying the exclusive emphasis on the party as the
DEMOCRACY AND REPRESENTATION
vehicle for representation; they were pursuing complementary (sometimes conflicting) ways of empowering women to make their needs better known.
The case for gender parity in politics should, I believe, be understood within this broader context, and to this extent, it confirms Hanna Pitkin's intuition. The argument for more 'descriptive' or 'mirror' representation does move in close parallel with arguments for a more participatory form of democracy; and those concerned with the under-representation of women in politics do look to additional mechanisms of consultation and accountability and participation that would complement our occasional vote. We do not need this additional ammunition to argue for more women in politics; there are arguments enough from justice or interests that provide a basis for substantial change. But as a more profound set of issues about democracy and representation, the case for gender parity is at its strongest when it is associated with the larger dream.
Notes
1. I use the term parity to indicate a rough equality between the proportion of women and men elected. My use of this term should not be confused with the arguments that have recently surfaced within the Council of Europe for so-called parity democracy. See Outshoorn (1993) for a critical review of this literature.
2. There are parties which operate quotas for youth—as with my own example of the 1972 Democratic National Convention—but when it comes to parliamentary candidatures, few people worry about the paucity of those under 25.
References
Chapman, lenny (1993), Politics, Feminism, and the Reformation of Gender
(London: Routledge). Grofman, B., Lijphart, A., McKay, R.B., Scarrow, H.A. (eds.) (1982),
Representation and Redisricting Issues (Lexington, Mass.: D. C. Heath
and Co,).
Kymlicka, Will (1993), 'Group Representation in Canadian Polities'. Paper prepared for IRPP project on 'Communities, the Charter and Interest Advocacy'.
Lovenduski, Joni and Norris, Pippa (1989), 'Selecting Women Candidates: Obstacles to the Feminisation of the House of Commons', European Journal of Political Research, 17,533-63.
Mohanty, Chandra (1993), 'Feminist Encounters: Locating the Politics of Experience', in Michele Barrett and Anne Phillips (eds.), Destabilizing Theory. Contemporary Feminist Debates (Cambridge: Polity Press), 74-92.
238
239
ANNE PHILLIPS
Outshoom, Joyce (1993), 'Parity Democracy: A Critical Look at a "New"
Strategy', paper prepared for workshop on 'Citizenship and Plurality",
European Consortium for Political Research, Leiden. Parry, Geraint, Moyser,  George, and  Day,  Neil  (1992), Political
Participation and Democracy in Britain (Cambridge: Cambridge
University Press).
Phillips, Anne (1991) Engendering Democracy (Cambridge: Polity Press). Phillips Griffiths, A. (I960), 'How Can One Person Represent Another?'
Aristotelian Society, Supplementary vol. xxxiv, 187-208. Pitkin, Hanna F. (1967), The Concept of Representation (Berkeley:
University of California Press). Ranney, Austin (1982), 'Comments on Representation Within the Political
Party System', in Grofman etal (1982), 193-7. Skjeie, Hege (1991), 'The Rhetoric of Difference: On Women's Inclusion
into Political Elites', Politics and Society, 19/2, 233-63. Verba, Sidney, Nie, Norman H., and Kim, Jae-on (1978), Participation and
Political  Equality.  A   Seven  National  Comparison (Cambridge:
Cambridge University Press).
240