225 Economic malaise in 1987 forced the social partners to the negotiation table. The attained agreement was dictated by the goal to regain Irish cost competitiveness. Subsequent accords repeatedly emphasised these goals, as wage moderation was traded for income tax reduction. In contrast, the issues of social justice were “treated as subsidiary to competitive imperatives” (Roche 2007: ibid.). Hence, as Nolan and Maître (2000: 341) note for Ireland, “centralised wage setting is clearly not enough in itself to limit the growth in earnings inequality.” 4.5.2 Social Implications Although earnings are influential in determining the levels of income inequality, they constitute only one of several income sources defining household disposable income. Disposable income dispersion is influenced by the distribution of additional market income sources, state transfers and income taxation (Nolan 2003: 134). In the case of Irish household disposable income dispersion, the level of income inequality derived from direct primary market income is mitigated to a certain degree. Primary market income positions based on educational differentials driven by TNCinduced technological change feed through to define the positions of households in terms of disposable income distribution. However, secondary market income and the state via social transfers and taxation increase market income advantages. Again education levels as well as employment status are decisive in determining the risk of experiencing poverty further demonstrating the segmentation of Irish society. Household Income Distribution When interpreting data on the distribution of disposable household income in Ireland, the increase in disposable incomes over the period under review should be kept in mind. As can be expected, the rise in aggregate disposable income translated into an increase in disposable household incomes during the 1990s. Accordingly, average weekly disposable income adjusted for household size rose by 44% from IR£ 129.45 in 1994 to IR£ 187.23 in 1998 (Layte et al. 2001: 13).167 The following table displays the decile shares of disposable income for Irish households following economic recovery and during economic expansion in the 1990s. 167 1/0.66/0.33 scale used to adjust household disposable incomes for equivalised size (Layte et al. 2001: ibid). 226 Table 15 Equivalised Disposable Irish Household Income, 1987-2000 (%) 1987 1994 1997 2000 a Bottom 3.2 3.8 2.9 3.2 20 4.8 4.8 4.6 4.5 30 5.8 5.3 5.2 5.5 40 6.5 6.0 5.8 6.9 50 7.4 7.0 7.3 8.0 60 8.7 8.6 8.8 9.3 70 10.2 10.5 10.5 10.8 80 12.3 12.7 13.1 12.7 90 15.2 15.7 16.1 15.6 Top 25.9 25.5 25.7 23.6 Gini 0.330 0.331 0.348 0.324 Equivalised disposable income is based on 1/0.66/0.33 scale. Sources: O’Reardon (2001: 117), a Nolan and Smeeding (2004: 16). It is evident that the distribution of disposable household income only changed mildly. The top income decile accounted for more than a quarter of disposable income for the majority of the period. The top-heavy distribution of household income is displayed in the development of the Gini coefficient, which rose slightly and peaked in 1997. Figures for the year 2000 indicate that inequality was reduced compared to 1987 levels. The reduction is visible in the lower share of income for the top 10% and in the increase of income for the lowest 10% of households. The displayed figures illustrate that the recorded increase in earnings dispersion did not feed through to produce greater inequality in disposable household incomes. Hence, the inequality in the distribution of disposable household income in Ireland only marginally declined towards the end of the 1990s and remained static for the majority of the period under review. Although the distribution of income was unequal, the top income brackets did not break away from middle-income positions in terms of their income proportions. Instead, the bottom income deciles were left behind, as their income growth did not match the increase of disposable incomes of middle and higher income groups (Nolan 2003: 137). Moreover, an alternative study by the OECD, which focused on the distribution of disposable household income amongst the population of working age,168 shows that the growth in income inequality during the 1990s is associated with lower income groups losing ground compared to the development of disposable income for the six middle-income deciles (Förster/Mira d’Ercole 2005: 64). The expansion of 168 The focus on the working age population makes possible a concentration on the effects of increased employment, which was a distinguishing factor of Irish economic expansion in the 1990s. 227 middle-income decile shares emphasises the nature of the FDI-led development strategy as a middle class project. The middle-income deciles were able to increase their share of earnings by almost four percent reaching a total share of 57%. In terms of total disposable income, the middle-income groups were able to attain a share of 56%, expanding their proportion by 3.5% between 1994 and 2000. This income growth was the highest within the OECD for this income group during the mid- 1990s (Förster/Mira d’Ercole 2005: ibid).169 The analysis of the distribution of disposable household income also illustrates that the strong increase in educational differentials, fuelling the dispersion of direct market income throughout the 1990s, was mitigated. The lower level of recorded disposable income inequality compared to direct market income inequality is mainly linked to a lower correlation of earnings with inequality (Nolan/Maître 2000: 343). Hence, the role of direct market income in explaining disposable household income inequality declined, while other explanatory factors have gained in importance. Income from self-employment has increasingly become more unequal (Nolan/Maître 2000: ibid.). Förster and Mira d’Ercole (2005: 64) found that although 48% of income from self-employment was accounted by the top 20% of Irish income earners, their share in self-employment income fell by eight percent between 1987 and 2000. However, the middle-income deciles increased their shares by 7.5% to attain a proportion of 45%. In contrast, the lowest 20% of earners had a share of six percent of income from self-employment, increasing marginally by one percent between 1987 and 2000. Unsurprisingly, income from this occupational category was the most unequally spread, demonstrating the large differences in forms of selfemployed work and remuneration of self-employment (Nolan/Maître 2000: 342). The most important inequality factor is the rising unequal distribution of secondary market income, namely capital income from rent, interest and dividend payments (Nolan/Maître 2000: 343). As a consequence of economic expansion during the 1990s, the profit share of national income has been steadily rising since 1987. Whilst before 1987 the wage to profit ratio of national income stood at 70 to 30, non-agricultural profit income equalled wages for the first time in 1999 (O’Hearn 2001: 187).170 Finally, secondary market incomes are concentrated in middle and higher income groups, who have been able to supplement their market incomes with capital income sources. This increasingly applies to pension income (O’Reardon 2001: 129). The middle-income deciles had the largest share of capital income in 2000, displaying a 169 In comparison, the disposable household income share of the top income quintile stood at 36.2% and fell by 3.2% between 1994 and 2000. The lowest income quintile had a disposable income share of 7.5% and experienced a marginal decrease of 0.2% (Förster/Mira d’Ercole 2005: ibid.). 170 Nevertheless, as shown by O’Reardon (2001: 127-128) the increase in capital income over the period has not contributed to an overall rise in disposable household income inequality. This is linked to the possibility of the underreporting of capital incomes in household surveys. Furthermore, a large share of profit income is retained by firms and, in the case of TNCs, profits are repatriated. 228 proportion of 48% and increased their proportions by more than 2% between 1987 and 2000. They were followed by the highest 20% of earners with a share of 45% of capital income. Accordingly, the lowest 20% of income groups had a share of just below seven percent (Förster/Mira d’Ercole 2005: 64). Nevertheless, the dispersion of primary and secondary market incomes has been mitigated. The single greatest mitigating effect was growing employment. With employment increasing by 10% throughout the 1990s and unemployment falling from over 16% in 1988 to under 6% in 1999, more people gained access to market incomes (O’Reardon 2001: 131). Increased labour force participation by women further mitigated disposable income inequality. Although the rise in female employment is associated with an increase in part-time and low paid jobs, the growth in precarious “flexible” work forms did not translate itself into growing disposable household income inequality. The majority of female employment led to an increase in double-income households, therefore, supplementing disposable household income and creating an equalising effect (Callan et al. 2001: 11). Furthermore, although more incomes were taxed due to increased employment (Callan/Nolan 1999: 184), progressivity and social transfers can only explain a small percentage (12%) of the reduction of direct market income inequality in 1997 (Nolan/Maître 2000: 339-340). Instead, enlarged employment opportunities are far more important in explaining the reduced impact of earnings dispersion on disposable household incomes, as increased employment spread direct market earnings more equally between households (Nolan/Maître 2000: 343). In contrast, the low level of inequality reduction through social transfers also explains why lower income groups lagged behind in the dispersion of disposable income throughout the 1990s. Even though unemployment dramatically declined, disposable income growth for those households remaining dependent on social welfare transfers fell behind the increases of average household incomes, as social transfers were not linked to the development of average living standards (Nolan 2003: 137, 140). Poverty and Exclusion Turning to poverty developments during the 1990s, increased employment reduced poverty substantially. Still, households dependent on social transfers were disproportionately subjected to poverty (Nolan 2003: 138). The table traces the development of deprivation and income poverty (mean and median as well as the official “consistent” poverty indicator´. The latter combines the concepts of income poverty and experienced material deprivation (Layte et al. 2004: 10). 229 Table 16 Development of Irish Poverty Indicators, 1994-2001 (%) 1994 1997 2001 Income Poverty a Mean per week 60% 30.4 30.1 27.5 50% 19.8 17.4 18.4 40% 5.2 6.3 8.1 Median per week 70% 26.7 29.0 29.3 60% 15.6 18.2 21.9 50% 16.0 8.6 12.9 Basic Deprivation a All Households 24.0 14.9 8.3 Consistently Poor Mean per week a 60% 15.1 9.7 5.2 50% 9.0 6.7 4.1 40% 2.4 3.1 2.5 Median per week 70% 14.5 b 10.7 c 4.9 a 60% 8.3 b 7.8 c 4.1 b 50% 3.5 b 5.2 c 2.9 b All disposable household weekly income figures are based on the Irish equivalised scale (1/0.66/0.33). Source: a Schröder (2005: 50, 56, 81), b Layte et al. (2004: 10), c Madden (2002: 27) It is evident that the official figures mask an increasing gap between the recorded level of income poverty and experienced deprivation of people living in low-income households. The proportion of the population experiencing income poverty has been growing especially after 1994. In contrast, the population share experiencing basic deprivation declined from 24% to 8.3%, while those experiencing secondary deprivation, i.e. having difficulties in participating in general welfare such as maintaining a car or going on holiday, have been increasing (Schröder 2005: 64).171 This divergence is especially evident when comparing the development of the official poverty indicator of the EU Commission, which is set at 60% of median incomes, with the official Irish poverty figure (70%), as highlighted in the table. Whilst income poverty has been steadily rising since 1994, growing by more than 171 For an overview of the measurement of Irish deprivation Layte et al. (2001: 30-53) and Schröder (2005: 42) as well as Nolan and Whelan (1996: 88). 230 6%, the share of the population living in consistent poverty fell by more than 10% between 1994 and 2001 (Schröder 2005: 80). This discrepancy results from the difficulty in measuring deprivation, as it is based on subjective expectations and perceptions, which change with the evolution of living standards across time. Hence, the indicators used to measure basic deprivation, which were set in 1997, were outdated by 2001 (Nolan 2003: 138). Furthermore, economic expansion resulted in the increase in middle-sized incomes, leading to the growth in median incomes to greatly outpace the enlargement of average incomes. Median household incomes rose by 97% compared to average household incomes rising by 75% between 1994 and 2001 (Layte et al. 2004: 4-5). Consequently, as displayed in the table, the income gaps measured in median income figures show a larger increase than the figures derived from average incomes (Schröder 2005: 51). Thus, although poverty was substantially reduced, the fact that more than a fifth of the Irish population had a disposable income of 60% of median earnings in 2001 also displays the high degree of income inequality. Moreover, as economic recovery turned into economic expansion after 1994, the gap between low and middle-income households increased, illustrating the combination of economic growth with inequality in Ireland. The following table displays the poverty rates in terms of the employment status in the 1990s. Table 17 Labour Market Status of Irish Poor, 1994-2001 (%) 1994 1997 2001 Share a Rate b Share a Rate b Share a Rate b Employed 9.1 3.2 16.1 3.6 19.0 1.8 Self-employed 3.2 4.6 3.8 3.9 1.1 0.5 Farming 3.3 5.5 0.8 1.0 2.0 1.3 Unemployed 45.0 52.2 39.0 42.7 14.3 19.1 Ill/Disabled 8.3 36.2 9.5 31.7 17.8 22.5 Retired 4.8 6.1 6.4 5.5 12.1 5.3 Home Duties 26.4 28.8 24.4 28.8 33.7 12.3 Active 60.6 12.2 59.7 8.4 36.4 2.4 Inactive 39.5 20.4 40.3 15.0 63.6 11.0 Total 100 14.5 100 10.7 100 4.9 a respective labour market status of the main household earner as a proportion of “consistently poor” households; b Official Irish poverty definition of deprived households with an equivalised disposable income of less than 70% of the median disposable income in terms of labour market status. Source: Schröder (2005: 59) 231 The impact of economic expansion and subsequent labour demand on Irish lowered overall poverty levels. However, when poverty rates are decomposed according to labour market status, households dependent on social welfare transfers are disproportionately classified as poor. Labour market status has become the most important factor defining poverty in Ireland. The share of the status of labour market activity as a poverty criterion fell from almost 61% to 36% between 1994 and 2001. In contrast the proportion of inactive poor households has increased. Whilst the poverty of rate of households with an employed household head has fallen by almost 50%, their share of poor households has risen by almost 10%. By 2001, almost a fifth of the poor were households headed by an employee, which relays the large increase in low paid employment. Furthermore, those households with a head unable to work due to illness or disability had the highest level of poverty, followed by the long-term unemployed and those working in the household. The latter accounted for a third of all poor households and includes a large proportion of single mothers, who recorded a poverty rate of 24% in 2001 (Schröder 2005: 60). Similarly, although child poverty has been significantly reduced during the 1990s from 15% in 1987 to 6.5% in 2001, the risk of an under 18 year-old to experience poverty increases with the number of children in the household as well as for children growing up in households headed by single earners (Schröder 2005: 61). Even though old-age poverty has been slightly reduced, falling from 6.1% in 1994 to 5.3% in 2001, their proportion of poor households has increased. However, the proportion of pensioner households experiencing deprivation is low compared to the strong increase in retired households experiencing income poverty with less than 70% of equivalised median income, which grew from 5% to 49% between 1994 and 2001 (Schröder 2005: 61). Again this strong increase illustrates that the growth of net incomes of lower income groups in general and of social transfer dependent households in particular did not match the large expansion of middle-income households. When further decomposing poverty rates in accordance with educational attainment levels and occupational categories, important dynamics pertaining to the level of skills and social class are revealed, displaying the labour market dynamics resulting from TNC-induced technological change. Households, whose head had no educational qualifications, recorded a 50% risk, those with intermediate levels 30% and the poverty exposure risk for those with accomplished secondary education stood at 20%. Furthermore, the risk of exposure to multiple years of poverty was higher for the low skilled than for the skilled and highly skilled in 1998 (Layte at al. 2001: 71-72). Hence, the higher the level of educational attainment, the lower is the risk of experiencing poverty. Closely related to these dynamics is also the issue of the origin of the social class of the poor. Layte et al. (2001: 73-75) as well as Nolan et al. (2000: 207) show that unskilled professions are disproportionately prone to experience poverty in comparison to other social classes. Unskilled manual workers are also more prone to endure 232 income poverty set at 60% of mean income with a 60% poverty risk in 1998 (Layte et al. 2001: 75). Again, members of the professional and managerial class had the lowest risk of poverty exposure with a 6% chance of experiencing poverty for more than one year measured as 50% of mean income (Layte et al. 2001: 73). For the self-employed the risk of poverty stood at 25%, displaying again the wide differences in forms of work and incomes from self-employment. In contrast, the poverty risk for unskilled manual workers was far higher with 40%. The latter also faced far longer rates of poverty than the other social classes (Layte et al. 2001: 74). The displayed class polarisations have also to be seen within the context of social mobility.172 Again the level of educational attainment in the Irish Republic is decisive in explaining the level and predicting the possibility of social mobility. Ireland has traditionally had a low level of social mobility with a large stratification along social class lines due to the impermeable nature of Irish class barriers (Breen et al. 1990: 20-53). The structural transformation of the workforce as a result of economic expansion has contributed to the perforation of upward mobility barriers. This has increased absolute upward mobility for lower social classes into the professional and managerial class (Whelan/Layte 2004: 90). In general, the importance of white collar and skilled manual occupational categories expanded significantly during the 1990s (Whelan/Layte 2004: 94). However, the chances for a member of the professional and managerial class to remain in that class were four times higher than for those originating from unskilled social classes (Whelan/Layte 2004: 95). Furthermore, Ireland has not become more meritocratic. Access to education continues to be defined by class membership. Members of skilled classes disproportionately access higher education institutions, allowing them to remain in their advantaged positions (Whelan/Layte 2004: 101). The increasing heterogeneity of higher class origin is instead the consequence of labour demand resulting from the transformation of the occupational structure. As a result, the number of positions available within the professional and managerial skilled labour classes increased (Whelan/Layte 2004: 90). The persistence of inequality of educational opportunity points to continued class differentiation in the Irish secondary and vocational school system in spite of successive reform efforts (Smyth/Hannan 2000: 117). Hence, despite the growth in absolute numbers of schoolchildren completing secondary education, the relative inequalities between different social classes persist. In the absence of effective state measures to regulate local school competition, middle and upper class pupils as well as the most academically able are concentrated 172 Social mobility is defined as the possibility of the member of one class to move into a higher socio-economic group (i.e. upward mobility) or into a lower socio-economic class (i.e. downward mobility). When measuring social mobility, it is important to distinguish between two concepts. Absolute mobility refers to the rate of overall mobility of individuals from one class to another. Relative mobility measures the comparative chances of an individual originating from a certain class to enter another class (Whelan 1999: 135-136) 233 in selective secondary schools. Vocational schools, whose curriculum was extended to provide secondary education in 1967, are disproportionately attended by children with working class backgrounds, many of them lower ability pupils coming from deprived and socially excluded backgrounds (Smyth/Hannan 2000: 113). Furthermore, the rising importance of the correlation between educational credentials, employment and incomes greatly disadvantages children with working class backgrounds, in particular those from unskilled households (NESC 2002: 232).173 Hence, increased absolute social mobility is not a result of increased state activism. Instead, it is a consequence of economic expansion, which increased the number of higher class occupational positions. Beyond the aggregate figures, educational inequalities based on class origin continue to persist. They define the chances of an individual to experience social mobility. Therefore, the lower the class origin, the lower is the level of education, the higher is the propensity to be unemployed or to accept under average earnings or to experience poverty (Whelan/Hannan 1999: 291). 4.5.3 Poverty, Inequality and the Irish State Benign observers, such as Schröder (2005: 82), are jubilant in their assessment of Ireland’s record of unemployment and poverty reduction. They claim that the Irish example shows how a lean welfare state with polices prioritising economic growth can successfully tackle social exclusion without massive redistribution. However, as demonstrated, the enlargement of national income was not distributed equally. Part of the growth in national income was exported in the form of profit repatriations by TNCs, prompting O’Hearn (2001: 187) to exclaim, “The main recipients of economic growth in Ireland were a foreign capitalist class rather than a domestic one”. Nevertheless, profit repatriations are the flipside of the FDI-led development strategy. More realistically, the level of failure to cater for an equal redistribution of the far larger share of national income remaining in the country has definitely benefited the domestic capitalist class. It, therefore, brings the role of the Irish state back into the forefront of the discussion, as redistribution outcomes are the result of political decision-making. Hence, the lack in state capacity to attempt to spread the fruits of economic growth more evenly is voluntary, resulting from political efficacy and policy legacies (Cousins 1995: 159). 173 Over the period of 1996-1998, early school leaving rates are the highest amongst children from unskilled households. 3.2% of the school leavers with no qualifications came from semiskilled households and 9.1% of non-qualified school leavers originated from unskilled households; the national average was 3.7%. The level of completed secondary and tertiary education was the lowest. 65.4% of pupils from the unskilled manual class and 75.1% of semi-skilled class pupils completed secondary education, while the national average was 80.6% (NESC 2002: 232).