Women and the Constitutional Politics of Care in Ireland

In 1937, Eamon de Valera, president of Ireland and founder of the country’s largest party, Fianna Fáil, introduced a draft of a proposed new constitution. Although not all Irish voters were convinced that this was necessary, seeing as the existing constitution was enacted only fifteen years prior, the section on the family quickly caught the attention of the public. In what came to be known as the “care clause” of the family section, Article 41.2 asserted that “the State recognizes that by her life within the home, woman gives to the State a support without which the common good cannot be achieved. The state shall, therefore, endeavour to ensure that mothers shall not be obliged by economic necessity to engage in labour to the neglect of their duties in the home.” Under Irish law, constitutional changes must be ratified by public referendum, and despite widespread protests from women’s groups, the new constitution overwhelmingly passed.

In March 2024, Irish voters returned to the ballot box to amend Article 41.2, in no small part because the “duties” language that seemed contentious from the perspective of the early twentieth century was viewed as irredeemably archaic by the twenty-first. The government led the charge to amend the constitution and proposed replacing the original text with the following: “The State recognises that the provision of care, by members of a family to one another by reason of the bonds that exist among them, gives to Society a support without which the common good cannot be achieved, and shall strive to support such provision.” Despite endorsements from almost all parties in the national legislature and highly visible civil society groups, the referendum failed by historic margins: 74% of participants in the referendum voted “no” on the amendment. Why? And what lessons are there for understanding the intersection of gender, constitutional law, deliberative democracy, and social policy?

Policy Origins

Article 41.2 as originally written made two key claims. The first, and perhaps the most controversial at the time, was the claim that the proper role for an Irish married woman was in the home. In making the case for Article 41.2, de Valera, the constitution’s primary architect, argued that his intent was to ensure that “the inadequate strength of women or the tender age of children should not be abused by the need to enter the workforce.” ¹ This assertion came to a surprise to those Irish women who had been actively involved in the independence movement of the preceding decades. In what might be considered a prototypical response from activist women of the era, Margaret Buckley, the president of Sinn Féin — de Valera’s former political party — fumed that Article 41.2 treated women “as if they were halfwits” (McCarthy 2024).

In addition to making assertions about the role of women, Article 41.2 also made an important claim about the state and its role in social protection. Scholars of social policy have noted that individuals and families seeking social protection are confronted with a “resource environment” consisting of the state, the market, nongovernmental organizations like churches or neighborhood associations, and family/kinship networks (Levitt et al. 2017; 2023). Under Article 41.2, the role of the state vis-à-vis care work was decidedly murky. At the time, some women’s groups expressed hope that constitutional recognition of women’s work in the home could form the basis of claims for state provision of social policies such as a family allowance and healthcare (Beaumont 1997). However, de Valera deliberately used the term “endeavor” in describing the role of the government in supporting care work, noting that anything more specific would place too much of a burden on the state. As such, the proposed constitution effectively made the point that although the role of women in providing care work was seen as integral to the state, the role of the state in supporting care provision was, at best, aspirational.

This lack of commitment from the government to support women’s work in the home stood in sharp contrast to its concerted efforts to constrict opportunities for women’s work outside of it, including a ban on married women in the civil service passed in 1924 and a marriage ban on primary schoolteachers in 1932 (Redmond, n.d.). To the extent that state social policies emerged during this time, they targeted male breadwinners via unemployment payments, child benefits (paid to fathers), and widows’ pensions in case of the male breadwinner’s death (Kelly 1999). In short, Article 41.2 simply put into constitutional terms what had already been set out in statutory law and social practice.

Shifting Context

Why then, decades after its ratification, did reform of Article 41.2 seem within reach? Reform efforts were facilitated by three key contextual changes that evolved over the last 30 years: the waning power of the Catholic Church, shifting public opinion about the role of the state in social protection, and the emergence of the Citizens’ Assembly, a deliberative body focused on constitutional reform.

Article 41.2 was deeply influenced by Catholic social teaching, which assumed heterosexual marriage as the anchoring principle of Irish social policy. This was in no small part due to the influence of Father John Charles McQuaid, the future archbishop of Dublin, who allegedly drafted the original version of Article 41 (McKinney 2022). However, as Catholicism’s grip on the Irish population weakened and an increasingly secular population mobilized, much of the social conservatism embedded in the constitution was excised or replaced (Green-Pedersen and Little 2021): Irish voters removed the ban on divorce in 1995, legalized same-sex marriage in 2015, and in 2018, removed the constitutional ban on abortion (which was initially put into place — also by referendum — in 1983).

In addition to social changes, Ireland underwent a significant economic change as well. Long plagued by high unemployment and emigration, economic reforms in the 1980s spurred a period of unprecedented growth. Although Ireland’s historic lack of state involvement in social welfare was in no small part due to its poverty, crying poor became a harder sell to the Irish public when Ireland’s economy converged with its wealthy Northern European neighbors yet social spending did not (Whelan 2019). Subsequently, Irish civil society demanded more state support for care work, rather than overreliance on families — that is, women (Hilliard 2019).

A final twist happened in the early 2010s when the process of constitutional reform was dramatically changed by the introduction of a citizens’ convention on the constitution, which eventually evolved into the Citizens’ Assemblies (CA). For each cycle of the CA, individuals are randomly selected to participate in a series of workshops related to a specific constitutional issue where they would hear competing expert views, deliberate among themselves, and then vote on proposals for reform (Citizensinformation.ie 2024). These proposals are compiled into a public report and often taken up by the national legislature; as such, Ireland has emerged as a global model for participatory governance (Courant 2021; Loughnane, Kelleher, and Edwards 2023). Like the referendums on same-sex marriage and abortion, the care amendment proposal was a result of this process: the 2020–2021 theme of the CA was “gender equality,” and Article 41 was a prominent component of its deliberations.

In their June 2021 final report, the CA recommended that Article 41.2 be removed entirely and be “replaced with language that is not gendered and obliges the State to take reasonable measures to support care within the home and wider community” (The Citizens’ Assembly 2021, 53). This represented a radical departure from the original text: by making it gender-neutral and changing the language about the role of the state from “endeavour” to “oblige,” the proposed framing would expand the resource environment for care work beyond family and kinship networks to envelop the state as well. The report also noted that even if gendered language were removed, gendered work remained, as 98% of carers in Ireland (paid and unpaid) were women, and among full-time unpaid carers, women outnumbered men 2:1 (2021, 59). As such, the CA also recommended more support from the state, including better wages and pension benefits, for the disproportionately female carer workforce.

Given a Catholic Church socially and politically weakened by multiple abuse scandals, the clear shift in voter sentiment, and the institutional backing of the CA, proponents of the 2024 initiative were optimistic that Article 41.2 could also be modernized. After successful passage of the care amendment through both houses of the Irish legislature, the government deliberately scheduled the constitutional referendum for March 8, 2024, coinciding with International Women’s Day. In the wake of recent successes with the referendum process, and the unpopularity of the original “duties in the home” language, it perhaps had reason to be optimistic. So why, then, did it fail?

Explaining Failure

The failure of the care amendment was certainly not due to widespread support for the original language. Government Minister Roderic O’Gorman opened the debate in Dáil Éireann, the lower house of the national legislature, by noting that this was an opportunity “to remove the archaic and sexist reference to a woman’s place in the home.” ² Indeed, many of the women representatives in the Dáil made reference to the fact that their presence on the floor contravened the intent of the original article; one quipped, “I assure the Chair that I have neglected my duties in the home today and probably every other day that I have been in the Dáil, and I am not really that bothered about that.” ³ Nevertheless, a review of legislative debate transcripts and Irish media coverage suggests two distinct lines of opposition that emerged from politicians and civil society groups that one might have expected to support amending Article 41.2: opposition based on the scope of the reform and the ambiguity around its impact.

The first wave of opposition to the proposed language was based on scope: the reform did not represent much of a departure from the status quo, for care providers or regarding the role of the state. The rewording did not, as the CA recommended, oblige the state to “take reasonable measures” when it came to care. Instead, the new wording seemed to swap de Valera’s “endeavour” for the equally mushy “strive.” Subsequently, a good number of voters were peeved that the government chose to strip down the suggested language from the CA. In addition, by limiting state obligations and keeping care work within the family, the government was, according to its critics, just trading one set of gendered language for another: in effect, the proposed language change swapped out de jure language about the role of women in the home for de facto language. Finally, disability activists were particularly angry about the continued emphasis on the family as a source of care. Their preference was to see the state play a greater role in supportive care services that could facilitate living independently of one’s family, and for there to be stronger legal and economic protections for carers (Wilson 2024).

Second, opposition to changing the existing language came from a sense that nobody really knew what the implications of the change would be. As one legislator noted, what a “YES” vote meant for the same-sex marriage or abortion rights referendum was clear, yet this was not the case for the care amendment. ⁴ Concerns about the implications of the language emerged from within the bureaucracy as well: according to press reports after the referendum vote, officials in the Department of Children pushed for creating a government obligation to support care work and, as part of their campaign, cited the CA report (McGee 2024). This raised red flags in other departments, concerned that a statutory requirement to support care work would end up being mediated by the courts rather than the legislature. These concerns were not misplaced: research on constitutionally embedded social rights in South Africa has found the tension between legislative and judicial responsibility over policy to be a key sticking point in implementation (Sunstein 2001).

Lessons Learned?

Materially, Article 41.2 has never had a significant positive impact on carers, and symbolically, the “duty” language is somewhat embarrassing but ultimately meaningless for a majority of voters. When all was said and done, the failure of the care amendment seemed to have had a much larger effect on the government than the care community: the Taoiseach (prime minister), Leo Varadkar, immediately resigned once the referendum results were clear. So, what can we take away from this? There are two lessons here with implications for cases beyond Ireland.

The first lesson for policy makers is to recognize the importance of de jure versus de facto policy, particularly when it comes to policies around gender and the private sphere. The government’s claim that it degendered the language of Article 41.2 belied the lived reality of the care sector, which is dominated by women. Although many were uncomfortable with the “duties in the home” phrasing of the original text, replacing “women” with “families” while eschewing any additional state responsibility for legal and material support for carers seemed like a pointless exercise. As one legal scholar noted, the country was “being asked to replace a non-operative clause containing outdated and patronizing language with a new non-operative clause with slightly more acceptable language” (Cahillane 2024). Perhaps it is not surprising then that the public said no.

The second key takeaway here has to do with what seems to be a growing crisis in Western liberal democracies: the persistent chasm between what citizens say they want and the policies they get from their elected officials. The inclusion of social rights in constitutions was originally meant to force governments to remain attuned to the preferences of those who might lack political power (Sunstein 2001). But as we saw in the case of Ireland’s economically invisible but socially crucial care workforce, there were profound differences between the recommendation of the CA and the proposal from the government. There was an even bigger gap between the enthusiasm of the political establishment for the amendment and actual voter support for it.

If, as democratic theorists claim, a democracy is by definition a system of government where the preferences of the public are weighted and translated into policy by elected officials (Dahl 1973; Mounk 2018), this persistent chasm between policy preference and outcomes can be dangerous. For the last decade, we have seen how an unwillingness or inability to address voter concerns about globalization, austerity, and crime have facilitated the rise of authoritarian populists around the world who are very committed to giving voters what they want, but less committed to the peskier components of liberal democracy like minority rights, judicial independence, or rule of law. If a system as transparently deliberative as Ireland’s still cannot guarantee the enactment of popular social policies, the rest of us may be in real trouble.