New dimensions of inequality in Northern Ireland, 1998–2020

ABSTRACT

To understand inequality in the post-Good Friday Agreement (post-GFA) period, we must understand how inequality in Northern Ireland has been subject to stressors long acknowledged in the international literature, though seldom considered in this context. These ‘new’ stressors include destandardization, retrenchment, and financialization, which have largely replaced the class and employment differentials that characterised 20th century inequality. We explore the specific experience of inequality in Northern Ireland since Good Friday. In doing so, we contextualise its experience by detailing important structural and institutional changes over this time, explaining how they work in the unique socioeconomic context of post-GFA Northern Ireland.

KEYWORDS:

• inequality
• income
• political economy
• Northern Ireland

Introduction

This paper argues that despite considerable progress reflected in falling ethno-religious inequalities since the 1970s, Northern Irish society remains troubled on several grounds. Whilst modern inequalities register on indicators such as poverty, deprivation, and unemployment, the current state of inequality in Northern Ireland is also sustained by a ‘neoliberal consensus’ model of government, adopted and amplified in the post-Good Friday Agreement (post-GFA) years. What was presented as an opportunity in the late 1990s to construct institutions robust to the tumult of global capitalism, given the unique vulnerabilities of the region at the time, has instead yielded structures that have sustained or amplified its destabilising tendencies. Little structural change has occurred in the Northern Irish economy since 1998, and by 2008 it was clear that the peace dividend of FDI-driven growth in investment and quality employment had not occurred (O’ Hearn, 2008). Instead, the Protestant-Catholic employment differential has been replaced by a low-skill, low-pay trap, affecting equally those of all affiliations (Mac Flynn, 2017), whilst temporary and precarious employment has increased, as has the share of those on low pay (Wilson, 2019). Whilst those in the lowest income quintile depend on transfers for over 50% of their total household income (Tinson et al., 2015), the enhanced conditions introduced by welfare reform and consolidation of entitlements with reductions in core rates have led to greater difficulties for those in the most vulnerable categories. Household indebtedness remains high, with debt-to-income ratios in the region higher than the UK average (McAuley & Flaherty, 2020), whilst almost half of all households have no savings whatsoever (Tinson et al., 2015). The underlying weakness of its private sector economic structure, based as it is on low-wage and low skill services work (Teague 2016), is exacerbated by the weakness of organized labour, with falling unionization, and low utilization of collective agreements in the private sector (MacFlynn, 2020).

The purpose of this paper is not to add to the already considerable evidence of slow progress towards greater equality in Northern Ireland. Instead, it examines trends in the post-GFA years in the context of international research and theory on drivers of inequality, in an era where ethno-religious divides in socio-economic outcomes such as employment have largely disappeared. In an age where occupation is less meaningful in determining one’s capacity to lead a flourishing life, we suggest that previous emphases on employment may distract from other processes more fundamental to understanding the region’s development post-GFA. We focus especially on three new stressors of inequality that have received much attention in the international literature, but are less rigorously theorized in the context of Northern Ireland: destandardization as the declining quality of work and weakened bargaining power of labour, financialization as rising debt dependence, and retrenchment as enhanced commodification of welfare. The first section offers context to the GFA by establishing the state of knowledge on class and inequality on the eve of its passage, as well as charting the decline of previously ‘durable’ binary inequalities of religion with respect to employment up to the Great Financial Crisis. Subsequent sections address each of the new stressors in turn, showing how the drivers of inequality post-GFA are better conceptualized along these lines. To this, we add empirical data from sources including the UK Household Longitudinal Survey, Continuous Household Survey, and several secondary sources. We conclude by discussing the implications of this theorization of inequality in the context of Brexit, and the limited prospects of achieving greater equality under reunification – in the absence of fundamental changes in Northern Ireland’s economic and institutional structures.

The decline of binary inequalities

Research on inequality in Northern Ireland has typically focused on categorical inequalities – namely differences in outcome between religious groups (Aunger, 1975; Breen, 2000, 2001, 2003; Rowland et al., 2021). The role of religion in explaining consistent differences in stratification outcomes for Catholics and Protestants was established empirically by Aunger (1975) using 1971 census employment tables. His analysis revealed considerable differences in the occupational class structures of Catholics and Protestants, with Protestant predominance in upper occupational classes, and Catholic predominance in lower. Industrial segregation favoured Protestants, with Catholic male representation in Engineering and Textiles at 15% and 25% respectively. These findings challenged the commonly-held view that there was merely ‘a limited tendency’ towards Protestant predominance in higher occupational classes (Budge and O’ Leary 1973 cited in Aunger, 1975, p. 2). More detailed studies of class structure and mobility arrived towards the end of the century from Breen and Devine (Breen, 2000, 2001, 2003; Breen & Devine, 1998). Using mobility survey data from 1973 and 1996, Breen (2000) found strong evidence for convergence in the class structures of Catholics and Protestants, with weakening importance for ascriptive criteria such as ethnic group membership in determining class position. This was explained in part by greater access to education disproportionately benefitting Catholic class mobility. Although convergence in the class structure was detected, analyses of the 1996 cohort showed that education did not yield meritocratic effects in Northern Ireland, whilst ethnicity and class origin group continued to play a role in determining class position (Breen, 2003).

Today, it is clear that these previously stable ethnic differences have all but disappeared in the wake of greater educational access, class mobility, and occupational desegregation facilitated by equality legislation, and changes in the sectoral composition of employment. Growth in sectors favouring Catholic employment such as public services and construction, combined with declines in sectors such as manufacturing and engineering with historical Protestant dominance, has led to a weakening of this once stubbornly persistent employment differential (Rowland et al., 2021). Cultural shifts in ethnic self-identification have also weakened the explanatory capacity of these once-stable binaries. Attitudinal research shows that although religious categories of ‘Catholic - Protestant’ continue to explain substantial differences in political preferences and attitudes (Evans & Tonge, 2013), the association of religious groups with nationalist and unionist identities respectively has weakened (Hayward & McManus, 2019; McNicholl et al., 2019). The share of those identifying as ‘neither’ (Nationalist nor Unionist) increased from 30% in 1999, to a peak of 50% in 2018–2019 (Hayward & McManus, 2019, pp. 141–142, Coulter at al 2021: 192). The extent to which this ‘neither’ group may constitute a distinct category is suggested by other compositional features. Whilst there is little variation by age or national identity, 61% of this group are female, 26% are educated to degree level, 40% are employed in professional occupations, and 42% had lived for some time outside of Northern Ireland (Hayward & McManus, 2019). Interview data points towards the consolidation of this group around a distinct ‘Northern Irish’ identity, rooted in notions of a common people and place, but with disparate political ambitions (McNicholl et al., 2019). This appears to register also in electoral preferences, with the ostensibly middle-ground Alliance Party doubling its seats between the 2017 and 2022 assembly elections (Garry et al., 2022). 30% of voters of ‘no religion’ expressed preference for Alliance in 2019, and their support draws from a mixture of working and middle class secular voters, and both main religious communities (Tonge, 2020).

Despite the persistently high vote share of the main partisan parties, developments such as these point to a considerable weakening of ethno-religious identity as a meaningful dimension of inequality. The merits of such categories as a basis for understanding inequality is mirrored in wider debates on inequality. Tilley’s (1999) model suggested that durable inequalities arose not from differences on continuous attributes such as income, but from their allocation to distinct social groups. These groups or categories such as race or gender are seen as more consequential and stable, as they confer disproportionate claims on value-producing resources on some members, to the exclusion of others.

The classic case of such durable categorical inequalities in Northern Ireland is the discrimination which occurred under the ‘’Stormont regime’ (Breen, 2000), conferring labour market and political advantages on Protestants relative to Catholics. This disadvantage was in turn differently experienced at local levels, where the unique sectarian geographies of Northern Ireland conspired to compound disadvantage amongst the urban working class (Shuttleworth & Green, 2009). Internationally, Tomaskovic-Devey et al. (2009) found that although class position mattered in explaining wage inequality, the effect of class was amplified by categories of education, race, and language group. Thus, whilst Tilley’s work underscores the importance of categories in explaining the persistence and reproduction of inequality, it also reminds us that such boundaries are historically constituted, and subject to change. If such categories no longer offer meaningful explanations of the reproduction of inequality, where else might we look?

As employment has become delinked from religion in Northern Ireland, there is growing recognition internationally of a delinking of occupation and reward and weakening of core working conditions, or ‘destandardization’. The process of destandardization internationally includes changes in the nature of work, such as weakened security of tenure and rising precarity, greater submission to performance monitoring, increased use of subcontracting and short-term contracts, and de-unionization (Kalleberg, 2009). This is often linked to new forms of work organization in leading companies, such as the ‘platform economy’ which has acted as an accelerant of precarity (Vallas & Schor, 2020). In the course of the Great Financial Crisis, Northern Irish involuntary non-standard part-time employment grew from 9.4% to 18.1%, the largest increase of all U.K. regions, and double the U.K. average (Green & Livanos, 2015, pp. 1226–1227). These phenomena are evident in professions previously thought immune to such changes, including law, medicine, and academia. In this context of high non-standard employment, where low income workers exhibit high dependence on transfers to maintain household income, and high levels of unsecured debt, underpinned by a state whose growth strategy remains wedded to accelerating these processes – irrespective of ethnonational identity (Coulter, 2014, 2019) – an alternative model is urgently needed. The following sections discuss the specific manifestation of these processes of destandardization, financialization, and retrenchment as new stressors of inequality.

Destandardization: sectoral change, precarity, and worker’s bargaining power

Discussing the nature and structure of work in Northern Ireland gives an opportunity to examine its economic context post-GFA, but especially, post-Great Financial Crisis (post-GFC). Northern Ireland’s economy differs markedly from its southern Irish and British counterparts in several respects. It commands neither substantial Foreign Direct Investment (FDI) that has proven so lucrative to the Irish state, nor does it possess financial sectors comparable to the capital cities of its neighbours. Its manufacturing base comprizes a small number of increasingly vulnerable heavy industries (those centred on the shipyards and aircraft manufacturers), whilst the remainder comprises Small-Medium Enterprises, lacking the capacity to capitalize on international supply chains. Placing current employment in comparative-historical context (see Table 1) shows long-term inertia in its economic structure, and little change in the pre- and post-2008 crisis years. Historically, the presence of a large public sector in Northern Ireland was key to mitigating the effects of industrial decline, and for maintaining effective demand as the economy stagnated during decades of conflict. This is reflected in the rapid proportional growth of public services employment from 26% to 37% (1974–1985), and declining manufacturing employment since 1974, as shown in Table 1. Before the crisis of 2008, public employment accounted for approximately 30% of total employment, and 60% of regional GDP, compared with 35% of GDP in the Republic of Ireland (Smyth & Cebulla, 2008, p. 182). Relative to other sectors, this dominance of services has shown little change. Historically, state buffers such as public employment, along with a 16% share of social housing in the total housing stock provided some degree of insulation from the vicissitudes of global capitalism.¹

Table 1. Percentage Employed in Key Sectors by NACE Industry Sector Codes (1974-2018). Source: Quarterly Employment Survey Historical Tables.⁹

	1974	1985	1996	2007	2018
Manufacturing (C)	32.8	20.8	17.6	11.7	11.28
Construction (F)	8.0	5.8	4.0	6.2	4.4
Retail (G)	12.3	13.5	15.3	17.4	16.9
Accomm. and Food Service (I)	1.8	2.9	4.9	6.0	6.6
Finance, Real Estate (K-L)	2.1	2.4	2.7	3.9	3.6
Scientific (M)	1	1.5	2.3	3.2	4.4
Admin Support Services (N)	1.6	2.1	3.5	5.4	7.2
Public services (O-Q)*	26.0	37.0	37.0	34.0	32.1
Arts, Ent, Rec. (R)	.9	1.5	1.8	1.9	2.0

*Category includes Administration, Education, and Health.

Despite this, the region has seen profound changes in the quality of employment resulting in a weakening connection between work and reward, and the dominance of service work (sectors G, I, N, R in Table 1) has a role to play in this. Since the GFC, investment and employment growth have tended towards low value-added, low wage service work, consolidating Northern Ireland’s position as a low-wage economy (Coulter, 2014, pp. 766–767). In 2019, 28% of Northern Irish employees earned below the ‘Real Living Wage’, a measure that adjusts nominal wage rates for costs of living (Wilson, 2019). While post-crisis unemployment recovered from 7.9% (Oct-Dec 2010) to 2.5% (Jul-Sept 2019), there remain serious concerns about the quality of work that has accompanied this ostensible recovery. Young people face particular challenges around employment de-standardization, with extensive use of short term and zero-hours contracts, unpaid internships, low pay, and competition for high-quality positions. Sub-contracting has also eroded employment security and pay in sectors once dominated by public employment such as housing, health and social care. For instance, the civil service spent £200mn in health and social care agency staff alone from 2010-2015, of an estimated economy-wide spend of £369mn (McVey, 2018, p. 46). Precarious employment rose in the post-crisis years by 25%, such that by 2016, 11% of the workforce were ‘self-employed without employees’. These workers were more likely to be below degree educated, male, and working in construction, finance, transport and communication (Irish Congress of Trade Unions, 2017).

Neither do workers in Northern Ireland widely benefit from either statutory or organizational labour market protections such as labour laws and union membership. This is important as the predominance of service work is linked internationally to sustained inequality, and poorer worker outcomes. Survey data estimates show that approximately 53% of workers in Northern Ireland work in jobs where pay and conditions are subject to negotiations between an employer and trade union (MacFlynn, 2020). This varies from 91% in public administration to 48% in manufacturing and 6% in construction (ibid: 10). Overall, Northern Ireland recorded the highest rate of all U.K. regions of employees whose pay is affected by collective agreements (44.2%, U.K. rate 26%), along with the highest recorded regional union density of 35%, compared to a U.K. rate of 23.4% (Department for Business, Energy & Industrial Strategy, 2018). The figure of 35% is conditioned here by the presence of disproportionately large public employment. The growth of service work thus does not bode well for future labour security, as it tends to make coordinated bargaining beyond firm level difficult (Flaherty & Ó Riain, 2020, p. 1044). The Northern Irish workforce also suffers from skill mismatching, and a particular inertia around upskilling. Its economy has been described as one of ‘low skill equilibrium’ (Mac Flynn, 2017), with high out-migration amongst skilled workers, predominance of low productivity small-medium enterprises, and high-turnover, low-skill service employment.

A key indicator of workers’ bargaining power is labour’s share of national income, or the percentage of national output (Gross Domestic Product, Gross National Income or Gross Value Added) going to workers are wages and salaries (Atkinson, 2009; Kristal, 2010, 2013). These trends bear consideration both in assessing the current state of labour security in Northern Ireland, and identifying potential threats under future constitutional scenarios. Northern Ireland’s labour share is below both the UK, and the EU average, which is unusual given its high share of public employment, and lack of high productivity sectors (Figure 1). Ireland’s labour share is currently the lowest in Europe, at 30.3% in 2021, relative to the EU-27 average of 55.5%. This erosion of worker’s share and of their collective bargaining power results from a combination of the dominance of finance and FDI sectors in Irish economic output, as well as weak union power, lack of centralized bargaining, and a residual welfare state (Flaherty & Ó Riain, 2020). Post-crisis (2010) labour share declined by 0.2% per year in Northern Ireland, and by over 2% per year in the Republic of Ireland (AMECO, 2022). With trade union coverage concentrated in public sector occupations, the potential erosion of public employment through falling recruitment and subcontracting (McVey, 2018) is likely to be especially damaging in Northern Ireland. Coupled with the data presented above on sectoral change, this bodes poorly for addressing the underlying issue of poor industrial growth capacity into higher value-added sectors, whilst protecting workers’ rights.

Figure 1. Comparative Wage Shares, 1960-2020. * Northern Ireland’s wage share is expressed as a percentage of Gross Value Added.⁷

Failure to address the fundamental weakness of labour in Northern Ireland does not augur well, considering the emphasis placed in reunification modelling exercises on multinational growth in the region. Simulations of the economic impact of reunification emphasize the potential payoffs of corporate tax harmonization between North and South (KLC Consulting, 2016). This is not a credible prospect for a state such as the Republic of Ireland, which poorly polices corporate tax compliance. The Irish State appealed a European ruling, which characterized its taxation of tech company Apple – and its underpayment of €13bn in corporate tax – as illegal state aid. Its reluctance to enforce regulation on key financial and corporate sectors, coupled with its ongoing erosion of public health provision and social housing, would be unlikely to address any of the structural and institutional deficits currently driving inequality in Northern Ireland. The underlying myth of multinational growth driving public wellbeing in Ireland is also revealed in its underwhelming comparative performance on healthcare, childcare, and education relative to per capita GDP (O’ Boyle & Allen, 2021). Together, these processes of destandardization present a fundamental challenge to maintaining not just the distribution of society-wide personal incomes, but also maintenance of just returns for workers and the state relative to their labour input.

Retrenchment: poverty and welfare reform

A contradictory finding in the cases of both Northern Ireland and the Republic of Ireland is that inequality as measured through Gini coefficients and income decile ratios has fallen consistently over the 2010s (Tinson et al., 2015). Northern Ireland’s Gini index is lower than the Republic of Ireland’s, and its redistributive intensity (percentage difference between market and net inequality) is 0.21. In both regions, the state plays a considerable role in redistributing market incomes through taxation and transfers, making the difference between market and net income inequality in Northern Ireland the lowest of the UK. This is unsurprising considering the high dependence of low income groups on transfers, where for the lowest income decile, over half of total household income derives from transfers (Tinson et al., 2015). Poverty rates also improved modestly in Northern Ireland from 2017 to 2020, in contrast to Britain. This was driven largely by improvement in post-crisis employment rates, housing affordability, and the mitigation of some of the harsher elements of welfare reform (i.e. caps, non-introduction of the controversial ‘bedroom tax’, and waiting periods) by the NI executive (Joseph Rowntree Foundation, 2022).

Yet, while headline figures are encouraging, much is hidden in the detail. While post-crisis unemployment recovered from 7.9% (Oct-Dec 2010) to 2.5% (Jul-Sept 2019), strong regional inequalities remain in the experience of ‘recovery’. Data on regional Claimant Counts shows a range of 0.4% (Cultra Ward, North Down) to 11.8% (East Ward, Strabane).² At Parliamentary Constituency level, areas such as Foyle recorded a pre-pandemic claimant count of 5.1%, and Belfast West 4.1%, above the national average of 2.6%. From 2020, pandemic-era stabilization and income replacement measures mitigated some of the worst excesses of the pandemic. The Coronavirus Job Retention Scheme (CJRS) kept unemployment rates low (Wilson, 2021), whilst recipients of these furlough payments saw a fall in income of 13%, compared to a fall of 40% for new Universal Credit claimants (Institute for Fiscal Studies 2020). With these measures gradually withdrawn, coupled with a new post-Covid cost of living crisis, the impact on Northern Ireland in particular could be dire. Thus, despite progress on poverty rates since 2008, projections show a potential rise of 67% in destitution in Northern Ireland, relative to a 10% projected increase in Britain (National Institute of Economic and Social Research, 2022, p. 35). Inflation hits poor households especially hard as they spend higher proportions of income on necessities, without wage growth or benefit increases to compensate (ibid: 30). Globally, real wage growth had already slowed by 2017 to its lowest since the GFC (International Labour Organization, 2018), despite increases in economic growth, making clear the role the state has to play in poverty reduction and income protection, rather than the labour market and employment alone. For Northern Ireland in particular, rates of deep poverty remain high, as although 18% of the total population live in poverty, two-thirds of these live in ‘deep poverty’ with incomes 50% below the national median (Joseph Rowntree Foundation, 2022).

Whilst, in many respects, this is merely a continuation of ongoing fragility reinforced by a weak underlying economy, poor employment quality, and residual welfare system welfare reform has played a key role in maintaining the vulnerability of Northern Irish households to economic shocks. This also compounds the effects of low income and poverty for those caught in the ‘poverty trap’. Where a stronger welfare state and labour movements play a role in policy-setting and regulation, the effects of economic shocks and stressors such as inflation, and declining employment security are mitigated (Bengtsson, 2014; Bengtsson & Ryner, 2015). Such mitigating factors have been consistently absent from Northern Ireland, and its neoliberal leaning is shown clearest in its approach to welfare reform. To British and Irish conservatives, the Great Financial Crisis provided an ideal opportunity to apply far-reaching cuts to public expenditure. This was enacted upon the Northern Irish welfare system most profoundly by a ‘neoliberal consensus’ of unionist and nationalist governing parties (Coulter, 2014, 2019). As part of the Stormont House Agreement in 2014 and ‘Fresh Start’ agreement in 2015, a condition of continued operation of the devolved parliament was the implementation of wide-ranging changes to the core benefits system. Whilst consolidating several entitlements such as the jobseeker’s allowance, housing benefit, and child tax credits into a single credit system, these reforms also proposed additional conditions and reductions in core replacement rates for vulnerable groups – those close to pension age, and those in receipt of disability allowance (Browne & Roantree, 2013).

Evidence shows the considerable hardship the introduction of Welfare Reform has wrought in Britain. With the temporary exception of Northern Ireland on some measures, caps on housing benefit proportional to house size were introduced for low-income private tenants, followed in 2013 by the controversial ‘bedroom tax’, where social housing tenants in properties deemed larger than required were penalised (Donnelley, 2019). Benefit caps were introduced in Northern Ireland in 2016, along with new conditions, such as benefit sanctions for categories of unemployed claimants who failed to participate in work trials. The replacement of Disability Living Allowance (DLA) with Personal Independence Payment (PIP) from 2015–2016 eliminated automatic entitlement, introducing the controversial ‘fit for work’ assessment process, with several media outlets reporting the disqualification of previously qualified DLA recipients (Hodgson, 2017). Disability-related income supports are especially essential in a society where PTSD and associated mental illness rates are high (Bunting et al., 2013), and where the intergenerational communication of troubles-related trauma within families had led to elevated suicide risks (McLafferty et al., 2016). The intergenerational importance of this is reflected in Northern Ireland’s especially worrying child poverty statistics, where almost 24% of children are living in poverty (Joseph Rowntree Foundation, 2022). As the Covid-era income protection scheme ends, workers in more vulnerable sectors such as retail and services stand to lose out most. The £20 per week cut to universal credit – already amongst the lowest payments of any advanced economy – is especially worrying. Considering the majority of Universal Credit recipients are in working households (MacFlynn, 2020), this will place further stress on those low-income working households whose excess income is now consumed by rising costs of living (Table 2).

Table 2. Debt to Income Ratio by Country (2015)*.

Country	Debt to Income Ratio (DTI, %)
Denmark	294.9
Rep. of Ireland	168.3
Canada	166.2
Northern Ireland	158.2**
United Kingdom	142.7
Japan	126.6
United States	105.6
France	100.6
Greece	99.8
Germany	85.4

Financialization: personal and unsecured debt

In the absence of a functioning welfare state and weak labour protections, how did low-income households maintain their expenditure in the post-GFA years? In the late 1990s, and in the run-up to the Great Financial Crisis, one of the key changes internationally – and indeed one of the precipitating factors of the crisis itself – was financialization. The concept covers several important changes in policy and regulation since the 1980s including, but not limited to: the rise of financial sector output as a share of total national economic output, increasing and diversified compensation for top earners and executives, orientation of firm management towards shareholders rather than employees, and a supplanting of real wage increases with household and individual debt dependency (Krippner, 2011; van der Zwan, 2014). As the share of income accruing to labour continued to fall across much of Europe in the late twentieth century, real wages stagnated in the face of rising costs of living (International Labour Organization, 2016, 2020). Later research would identify the shifting dependence of economies such as the U.K. towards finance and financial sector output, as a contributor to rising inequality (Flaherty, 2015; Kus, 2012; Stockhamer, 2017), and a slowdown in economic growth (Tomaskovic-Devey et al., 2015). Debt was central to this process. As the cost of living rose in the face of stagnant wages, household debt became an important stopgap in meeting everyday expenses (Stockhammer 2008). As U.K. consumption rates remained stable over the years preceding the financial crisis, increased debt was used to sustain household cash flow (Stockhammer 2008, pp. 189–190).

Northern Ireland was not immune to the structural processes of financialization, and whilst rising property prices and sub-prime lending did not set as disastrous a context for the crisis as was the case in the U.S. and Republic of Ireland, levels of personal unsecured debt become problematic between the passage of the Good Friday Agreement in 1998, and the Financial Crisis of 2008. The composition of this debt matters, insofar as unsecured debt differs from other kinds of debts (such as mortgages and car loans) that are secured against physical assets. High-interest payday loan companies often market stopgap credit to vulnerable and low-income groups unable to meet high repayments due to compound interest rates (Ali et al., 2015), whilst the decision to access such products is often driven by existential considerations, rather than economic rationality (Brown & Woodruffe-Brown, 2015). Of the almost 600,000 low-income households in Northern Ireland in 2016, over 400,000 were found to use credit of some variety, with 50,000 of these using high-cost payday loans (Ellison et al., 2016, p. 29). It is difficult to conclude, in the absence of reliable data, if household credit decisions were shaped by aggressive marketing policies by banks, or by improved access to credit due to rising house prices prior to 2008 (Stockhammer 2008, p. 188), but, in Northern Ireland, both have likely played a role. In the two-year period prior to the Great Financial Crisis, the ratio of median property prices to median annual income rose from 5.2–9.2³, suggesting that households may have had access to additional credit via equity release.

Measures of total debt quantities alone are insufficient to assess the contribution of debt to inequality, however. In assessing the extent of problematic indebtedness, it is important to consider the experience of debt from the perspective of households. Household debt studies have typically worked with fixed-income thresholds to define problematic debt, where a household is considered in ‘problematic debt’ if its debt to income ratio exceeds 25% (Kempson, 2002). Another approach is to factor the ability of higher income households to service higher shares of debt to total income. This allows us to correct for the potential under-estimation of over-indebtedness amongst low income households. Using data from the Northern Irish Continuous Household Survey (CHS)⁴ 2007 debt module, we can account for this by adjusting the over-indebtedness threshold for different income quintiles. This gives us a comprehensive overview of the extent of unsecured debt in Northern Ireland on the eve of the financial crisis of 2008. This approach shows one-fifth of low-income households in Northern Ireland of this time were classified as over-indebted, with average monthly payments of £234 – bearing in mind that this calculation is specific to unsecured debt, which is most commonly used to finance ongoing household expenses (see Table 3). Immediately post-crisis, CAB NI data (2009) data showed that 33% of their clients owed less than £5,000, but that this amount was enough for those households to find their debt repayments problematic, as the cost of servicing that debt was high compared to their earnings. This is important for households where payments on smaller debt amounts carry larger interest rates requiring a greater proportion of their disposable income, potentially making those payments more of a burden (Manktelow, 2011).

Table 3. Pre-crisis over-indebtedness thresholds by Income Quintile (2007).

Income Band	Over-Indebted DSI (%) Levels^10	Median Monthly Income (GBP)	Monthly Repayments DSI > 25% (GBP)^11	% households over-indebted^12
Quintile 1	20	937	234	20.9%
Quintile 2	25	1583	396	12.4%
Quintile 3	30	2144	536	10.3%
Quintile 4	35	2919	730	10.3%
Quintile 5	40	5433	1358	10.7%

*‘Debt’ is household total debt including mortgages, as percentage of annual net household income.

**Author’s own calculations using British Banker’s Association data for total household debt, NINIS total number of NI households and NISRA average disposable household income.

We also find that several factors implicated in explaining poverty and deprivation are also associated with problem indebtedness. Lone-parent, families with dependents, sub-degree educated, and low-income households are most at risk, as are those where at least one member is working part-time. Odds of over-indebtedness are substantially higher for those households with multiple credit facilities and loans, with a little evidence that mortgages are important (McAuley & Flaherty, 2020). Knowing that the consumption of ‘problem’ credit instruments such as high-interest bridging loans is likely more prevalent amongst low-income households, coupled with Northern Ireland’s lower savings rates, this further affirms that already-vulnerable households are more likely to have borne the worst consequences of debt. Northern Ireland is uniquely vulnerable relative to Britain in this respect. As unsecured debt is marketed principally to low-income households, such short-term credit instruments work only insofar as their consumers can rely on a regular flow of income to meet repayments. Evidence instead shows a cyclicality to credit consumption amongst low-income households and social tenants, who may use overdrafts to resolve credit card debt, which ultimately cannot be paid down (Ellison et al., 2016, p. 38). Compound interest means that the rate of increase in total repayment liabilities is non-linear, and households that opt for small bridging loans can quickly find themselves in problem debt.

The situation of households has improved little since the crisis. Equity release has evaporated as a source of bridging credit, and the tightening of mortgage lending rules has likely exacerbated inequalities in wealth formation due to the effective lockout of younger families and individuals from homeownership. OECD figures in Table 2 show a comparison of debt-to-income ratios (DTIs) across countries. In 2015, the UK had a total household debt to disposable income ratio of 142.7%, with NI households at a higher DTI (158.2%) than the rest of the UK, closer to the figure of the Republic of Ireland (168.3%). This figure likely reflects the effects of the NI housing boom and bust and the effects of post-conflict economic optimism on total debt to income levels, as well as lower disposable income levels in Northern Ireland than other regions of the UK (NISRA Well-Being Report, 2016).

U.K. Household Longitudinal Survey data from 2018 shows that although equal proportions of households report a subjectively comfortable financial situation in both NI and Britain (approximately 33%), the self-reported savings rate is low, with 36% of NI respondents reporting saving, compared to 39% in Britain. When decomposed by social class, 63% of professionals in Northern Ireland report having a savings account, but only 39% of routine workers. Thus whilst savings rates are marginally lower in Northern Ireland, when combined with higher burdens of debt and employment precarity, a pattern of financial precarity is likely to have exposed low-income households in particular to increased risk of economic stress. Whilst some of this was offset by stronger labour-market performance post-2015 (Joseph Rowntree Foundation, 2022), it was insufficient in addressing persistent long-term poverty. With high levels of income and employment precarity and extensive welfare retrenchment post-2008 as part of a wider British programme of fiscal austerity, it was a prime context for unsecured borrowing to proliferate, particularly as regulation of credit markets remained weak. Even without accounting for differences in base income levels across Northern Ireland and Britain, rates of debt remain higher for Northern Irish individuals within class groups. Northern Irish households in social class groups 3–5 of the NS-SEC are more indebted than their British counterparts, and this is especially concerning since individuals in these socioeconomic groups are most likely to suffer the effects of employment destandardization.

Exploring relative income differences and the declining significance of religion

This section offers a brief analysis of the impact of some of the above factors on relative income levels. This exercise might be approached in several ways, bearing in mind the limits of using absolute income levels as a dependent variable. Another approach might consider the odds of class membership, or membership of income quintile as a function of each of the below conditions, under more appropriate statistical control. The following graphs show conditional medians only and should not be interpreted in a regression context. Figures are generated using self-reported household income data from the 2016–2017 wave (8) of the U.K. Household Longitudinal Survey (UKHLS), which includes a sample of 1382 Northern Irish households, and 2550 individuals. Data were obtained from the UK Data Archive under special license access (University of Essex, Institute for Social and Economic Research, 2022). Figure 2 gives relative income differences for Northern Ireland only, whilst Figure 3 reports medians for the remainder of the U.K (England, Scotland, Wales). With reference to past studies that show persistence and decline in relative income differences along religious lines (Aunger, 1975; Breen, 2000, 2001, 2003), we observe how relative income gaps are now most pronounced along other dimensions, and the religious difference has all but disappeared. Granted, this is merely in terms of inequality of outcome (household income) and does not account for others such as persistent worklessness or unemployment differentials, although these have also declined into insignificance (Rowland et al., 2021). As such, this exercise merely points towards a likely avenue for future productive research and modelling in a more advanced multivariate context.

Figure 2. Median Income Difference by Category, Northern Ireland (2017). Source: U.K. Household Longitudinal Survey (Understanding Society).⁸

Figure 3. Median Income Difference by Category, United Kingdom (2017). Source: U.K. Household Longitudinal Survey (Understanding Society).

Figure 2 underscores the extent to which religion has declined as a predictor of relative income differences. This marks a continuation of the convergence in the class profiles of ethno-religious blocs evident from the late 1990s, when it was found that risk of low income for Catholics had dropped 38%, and risen 25% for Protestants (Smyth & Cebulla, 2008, p. 185). Some clues to the source and salience of the ‘new’ stressors of inequality are evident by comparing the income gaps produced by different variables, particularly those associated with employment destandardization, and financialization. Comparing the U.K. (Figure 3) to Northern Ireland, we find further differences in the income penalties experienced by different groups in both territories.⁵ The gender and benefits pay gaps are smaller in Northern Ireland, however, and a full explanation requires additional controls for tenure, and working hours, though some of this is likely driven by poor conditions prevailing in the service sector, irrespective of gender. Whilst both Northern Ireland and the U.K. show a non-linear distribution of income across NS-SEC class categories, small employers and own-account workers are second to professionals in relative income rank. This is unsurprising, given that Northern Ireland’s economic base is largely comprised of loosely-networked and locally-embedded SMEs, coupled with its sustained industrial decline which has relegated the income premium previously accruing to technical occupations. The opposite prevails in the wider UK, where category 4 (lower supervisory and technical) are below category 1 in terms of relative income, suggesting again that Northern Ireland has been unable to appropriately capitalisze on jobs that might be expected to spin out from strong tech sectors. The impact of destandardization on income is much more severe in Northern Ireland however. We see this most clearly in the difference between those in permanent and non-permanent employment, and those on zero-hours contracts. The monthly net income penalty for those in non-permanent employment in the U.K. is £340, and for those on zero-hours contracts, £279. In Northern Ireland, the non-permanent penalty is £448, and £525 for zero-hours contract workers.⁶

As the UKHLS lacks measures that would permit estimation of DSI, as with the CHS in the previous section, we instead look to the role of investments, thus capturing the wealth inequality aspect of financialization where certain households may enjoy an income or wealth premium by virtue of holding assets such as investments, stocks, securities, or high-interest savings. The financialization premium is stronger in Northern Ireland relative to the U.K., suggesting potentially greater wealth inequality in the region. Caution is needed here again, as the ability to realize flow income from investments is less certain than with earned income, though doubtless such items translate into a greater degree of income security over the medium to long term (Kus, 2016). In a context of extensive regional wage poverty (Wilson, 2019), the observed relative income gaps between those in standard and non-standard work should play a strong contributory role in rising inequality. As noted by others (McVey, 2018, p. 46; Mac Flynn, 2017; O’ Hearn, 2008), Northern Ireland suffers from both an equilibrium of low-skills, outmigration producing a ‘brain drain’, and historical lack of FDI. These structural issues are important in several respects. Research on multinational ‘greenfielding’ (Lamare et al., 2009) has shown that in countries with weak labour market protection, multinationals are better able to mobilize HR practices geared towards avoiding union recognition and engagement. The recent administrations of Harland and Wolff, and Wrightbus, coupled with ongoing supply chain threats against the wider manufacturing sector (i.e. Bombardier), have highlighted both the precarity of its ‘high value added’ industrial base, the lack of local strategizing around industrial upscaling, and employment creation. In the absence of growth prospects beyond low-wage service work, and little scope for local government intervention in labour market regulation which might protect employment standards, destandardization in the form of rising precarity, falling worker bargaining power, and increased nonstandard contractual security, is likely to be a key mechanism in continued growth in inequality.

Conclusion

We set out in this paper to argue that Northern Ireland has been left out of wider international research on drivers of inequality in the pre- and post-GFC years. Much of this work internationally focused on the roles of financialization, welfare retrenchment, and destandardization. By contrast, much work on Northern Ireland has focused either on relative differences in terms of ethno-national identity, employment structure, of challenges to industrial upscaling. We have here offered some preliminary evidence of the roles of these stressors in the unique regional context of Northern Ireland. We argued that to explain its persistent difficulties on several measures such as income differentials, quality of work, and long-term poverty, we must look to how these factors evolved and worked in the regions post-GFA, to the Great Financial Crisis, to the end of the last decade. Pending further research on the above factors in a quantitative context, we assert that these are indeed salient factors, and that the region offers much to substantiate international evidence on how these factors work to enhance inequality and otherwise pressure quality of life downward. Whilst we have not discussed this in a context of overall falling inequality as reflected in Northern Ireland’s positive progress on Gini scores, we assert that this is merely one measure of what is inherently a multidimensional phenomenon (income inequality), one that requires several indicators in order to by fully apprehended. From this expanded consideration, we conclude that workers appear to be extracting a declining share of their productivity from the Northern Irish economy, in a context of falling employment quality, persistent low-skill, low-pay work in the absence of a welfare state offering significant income replacement, and with private credit markets picking up the slack in terms of household expenditure.

What are the future prospects as the region enters a period of constitutional uncertainty? The economic policy challenges of Northern Ireland are often portrayed as a choice between the ‘high road’ approach of stimulating indigenous industrial upscaling and innovation, or the ‘low road’ of incentivising foreign investment through corporate tax reductions (KLC Consulting, 2016). Neither approach now seems appropriate. Whilst pockets of industrial employment remain in sectors such as aviation, these are precarious as parent companies face restructuring pressures both through international production cost and wage competition, and internally from shareholders seeking labour cost reductions. We were reminded of this precarity by the rescue of the Harland and Wolff Shipyard in 2019. As of 2019, progress on long-term industrial policy development had stalled in the absence of a functioning devolved government, and little interest from Westminster as Britain absorbed itself in several political crises related to Brexit. The prospects of future economic stability in the region are thus poor, with persistent difficulties over the implementation of the Northern Ireland protocol, which if repealed, could plunge the region into further economic depression on the back of a growing cost of living crisis, driven by rising inflation. Given the noted equilibrium between low skills and domestic output (Mac Flynn, 2017), coupled with its history of stunted high value-added investment, it seems likely that sectoral and employment-related processes will play the greater role in determining the shape of inequality into the future, with employment precarity and poor regulation driving greater credit consumption.

Socioeconomic change does not occur in a vacuum, however. State, industry, and labour play their respective roles in policy-setting and enacting structural change. This is key to understanding how socioeconomic context translates into real impacts on the distribution of income and economic security via industrial and labour policy, financial (de)regulation, and public spending pathways. The international literature has long established that economic growth alone is insufficient to generate positive outcomes for citizens. Whilst economic growth may raise national wealth and household incomes, in the absence of redistributive effort from the state and social actors such as labour movements, it will enhance rather than mitigate income inequality (Flaherty, 2015). This is shown to be especially so in the most recent phase of finance-driven growth, where increases in productivity were not evenly distributed to labour (Guschanski & Onaran, 2017; Stockhamer, 2017). We see the same tendencies repeated in Northern Ireland, especially so since the Great Financial Crisis of 2008. What of the prospects of addressing this politically? Demographic change has also seen increased the representation of both nationalists and ‘others’ at all levels of government and pushed the question of North–South reunification back to the medium-term political agenda. It is not clear that unification would yield any immediate economic windfalls, however. Northern Ireland is dependent on heavy subventions from Westminster, which still retains legislative control over key fiscal issues such as core taxation rates. The prospects of carving a viable way forward with the merger of two regimes as they currently exist is a poor basis for improvement in real standards of living and quality of life in Northern Ireland. And whilst the Republic of Ireland differs in many respects – its welfare state is equally residual with a low replacement rate, it has comparatively low investment in social and affordable housing, and an extensive system of private healthcare, operating parallel to a strained public system. Whether Northern Ireland’s short-term fortunes are secured by retention of the protocol or not, it is clear that it stands to lose, as much as gain, under any reunification scenario, especially in the absence of appropriate protections for workers.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Eoin Flaherty

Eoin Flaherty is Assistant Professor of Sociology at Maynooth University.

Martina McAuley

Martina McAuley completed a PhD in Sociology at Queen’s University Belfast on determinants of personal an household debt, and has since worked in a range of Non-Governmental Organisations based in Northern Ireland.

Notes

1 https://www.communities-ni.gov.uk/system/files/publications/communities/ni-housing-stats-17-18-full-copy.pdf

2 The Claimant Count includes data counting those on Jobseeker’s Allowance (JSA) and out-of-work Universal Credit Claimants. With the rollout of Universal Credit over this time, refinement of these mixed counts is ongoing. Full data are available at the following link:

https://www.nisra.gov.uk/statistics/labour-market-and-social-welfare/claimant-count

3 https://www.finance-ni.gov.uk/sites/default/files/publications/dfp/NI%20House%20Price%20Index%20statistics%20report%20Quarter%203%202019.pdf

4 The Continuous Household Survey (CHS) contained a debt module in its 2007 edition, and is the only source of information on unsecured debt levels in Northern Ireland. The CHS is one of the largest continuous surveys carried out in Northern Ireland each year by the Northern Ireland Statistics and Research Agency and has been running since 1983. The survey is a repeated annual cross-sectional study, and uses a multi-stage, stratified random sample with face-to-face interviews. Each year approximately 4,500 addresses are selected randomly from the Valuation and Lands Agency list of addresses to be contacted, and all members of the household aged 16 years and over were interviewed but only those aged 18 and over were included in the analysis.

5 Figures 1 and 2 do not report partial effects. These graphs display difference in medians within and between categories and territories, as indicated by category labels.

6 A full multivariate analysis with controls for education, age, gender, and sector is needed to corroborate this.

7 Devolved wage share data for Northern Ireland is not available from AMECO. Northern Ireland’s wage share is calculated as the percentage of gross value added accruing to waged and salaried workers (compensation of employees). This underestimates the share of compensation accruing to the self-employed, and the allocation of mixed income arising from self-employment, which may be attributable either to capital or labour (Sidhu and Dunn, 2018). The Northern Irish data displayed in figure 1 should not be taken as a correct ratio, but may be used cautiously to discern and compare trends.

8 The five-class National Statistics Socio-economic Classification (NS-SEC) scheme is used as a measure of social class, and numbers 1–5 denote the various classes. It is derived from occupation, employment status (employer, employee, self-employed), organisation size, and supervisory status. Such schemes typically focus on ‘attenuation of service relationship’ as a key dimension in distinguishing groups. Bearing in mind that Weberian class groupings such as this have been extensively critiqued, a more comprehensive analysis might focus on an inductive generation of new stratification groups using a wider range of variables. This framework does, however, permit comparison with those studies for which occupationally-derived class was salient. The five groups are categorised as follows: (1) Managerial and professional occupations, (2) Intermediate occupations, (3) Small employers and own account workers, (4) Lower supervisory and technical occupations, (5) Semi-routine and routine occupations.

9 https://www.nisra.gov.uk/publications/quarterly-employment-survey-historical-tables-june-2019

10 The Debt Servicing Index (DSI) is the percentage of monthly income going to debt payments, rather than the ratio of total debt to total income. This better captures the impact of debt on daily household expenditure. Using a sliding scale accounts for the differing ability of high income groups to service greater volumes of debt, and as such, as define problem debt as a DSI of >20% for quintile 1, 25% for quintile 2, etc.

11 This column shows the average repayments for those with a DSI>25% in each category.

12 This column shows the percentage of households defined as over-indebted using the sliding scale of DSI (see footnote 7).