The Survey on Income and Living Conditions, which forms the basis for most of our discussion on income distribution in Ireland, excludes key sources of income at the top and overstates the income of many low-wage workers. When we hear, ‘we have the most progressive income tax system in the world’, ‘the state does the heavy lifting’, and ‘Irish income inequality has been remarkably steady for a long time’, these narratives are based on this incomplete data.
The Survey on Income and Living Conditions, which forms the basis for most of our discussion on income distribution in Ireland, excludes key sources of income at the top and overstates the income of many low-wage workers.
A footnote on the issues around non-response and under-reporting for high-income earners in surveys will almost certainly appear in any research on income distribution.
‘EU-SILC (and household sample surveys in general) may not be the best source to study the tails of income distributions; other alternatives, such as total income statistics or taxation statistics, where available, may prove more useful e.g. in studies of “top incomes”, including property income, capital gains, stock options, or CEO pay.’
SILC data does not capture the income of Ireland’s 17 billionaires (the fifth highest per capita rate of billionaires in the world), for instance. Economists attempt to address this issue (with a certain amount of guesswork) with a weighting system to factor this in to estimates at the top of the distribution.
There are additional issues however around gaps in survey data related to important sources of income.
From 2020, capital gains will be included in income data from the Survey on Income and Living Conditions (SILC) with a catch-all question on additional income sources. To date, questions on capital gains (profit from the sale of an asset, such as a house) have not been included in the survey and as such, are excluded from estimates on income and income inequality. This data is the basis for various studies on income distribution in Ireland and frames the debate in an Irish context.
When we hear, ‘we have the most progressive income tax system in the world’, ‘the state does the heavy lifting’, and ‘Irish income inequality has been remarkably steady for a long time’, these narratives are based on this incomplete data (it’s also important to remember that these types of statements refer to income inequality only, not wealth inequality which is far more unevenly distributed in Ireland and arguably more important).
Recent literature highlights the importance of capital income in measuring the income of top earners and income inequality over the long run. In the US at least, capital income is heavily concentrated at the top of the income distribution, with roughly 75 percent of the benefit of the preferential rates on long-term capital gains and qualified dividends accruing to the top 1 percent of households. Capital gains in income are particularly important for examining the top of the income distribution and for analysis of income distribution over time. Between 1996 and 2006 in the US, changes in capital gains and dividends were the largest contributor to the increase in income inequality (SILC data includes dividends in income estimates). Figure 2 shows Eurostat estimates comparing GINI coefficients (a summary indicator of income inequality) in a selection of Nordic countries at the turn of the century, with and without realised capital gains. In Sweden in 2000, the difference in the two GINI scores was over five percentage points (for context, this represents a comparable difference in income inequality between Ireland (32.8) and the Nordic social democracies; Denmark (28.7), Finland (27.4), Norway (27.0) and Sweden (28.8)). The OECD estimate that the income share of the top 1 percent increased from 7 to 9 percent in Sweden when capital gains were included (2012). (They also point out that removing tax reliefs on capital gains would lead to more equitable outcomes in income, promote growth and reduce opportunities for tax avoidance for high earners.)
In Ireland, the revenue commissioners collected over €160 million or 20% more in capital gains tax in 2018 than in 2017 (1.02 billion compared to 841 million), compared to a 6.6 percent increase in income tax receipts. The exchequer only received €367 million in 2013. With a headline Capital Gains rate of 33% (after all the exemptions, of which there are many), this means at the very least €3 billion in income, mostly at the top of the distribution, is not included in survey data. The overall figure is likely considerably higher. For some scale, total employee earnings in the Irish economy were approximately €71 billion in 2018.
Issues further down the income distribution are less to do with incomplete data, but to do with how that income data is counted and presented in surveys.
Most studies on income distribution focus on dwellings (referred to as households in most surveys, including in SILC data) as the unit of analysis. For various reasons, this has failed to capture the changes in income for two relatively low-income groups of individuals over the past decade and more, sharing renters and adult children living at home, both of whom have grown significantly as a share of the population in recent years.
Twenty percent of the Irish population lived in rented accommodation in 1991. By 2016, this figure had increased to 30 per cent. The number of adults in their thirties living at home fell from 70,707 to 65,693 between 2002 and 2006 but increased to 91,000 by 2011, closely linked to economic conditions. Living at home past a certain age is usually a function of inadequate income, issues at the very centre of income inequality research. The latest census showed that 460,000 adults were still living at home in 2016. This comes to about 14 percent of the adult population. Between 2011 and 2016, although the population of 18-34 year olds fell by 11.3 percent, the share living at home with parents increased from 43.8 to 53.4 percent (the latest figure is 51.8 percent for 2018). The share of the same group working full-time but living at home increased by nearly 50 percent between 2011 and 2016 (20.8 to 29.8 percent). By 2018, this figure had increased to over one third (33.6 percent). As average rents have continued to outpace wage increased since 2018, the share of adult children living at home has likely risen in the meantime.
‘Households’ in CSO data (and elsewhere), are not really ‘households’ as we understand them. They would be more accurately described as dwellings and the income of family homes are counted the same as sharing renters i.e. income from individuals in the dwelling (a husband and wife for instance or strangers sharing rental accommodation, like a minimum wage cleaner and student) are all added up together to calculate ‘household’ income. The CSO have recently announced plans to address this issue. In future, surveys will have to establish a relationship beyond simply sharing living space to identify real ‘households’ as we understand them.
In addition, the income of the ‘household’ is equivalised to measure income inequality. The ‘at-risk-of-poverty’ threshold is set at 60 percent of the country’s median equivalised disposable income, for instance. The OECD provide a note on the indicator: ‘The needs of a household grow with each additional member but – due to economies of scale in consumption– not in a proportional way. Needs for housing space, electricity, etc. will not be three times as high for a household with three members than for a single person. With the help of equivalence scales each household type in the population is assigned a value in proportion to its needs.1 The factors commonly taken into account to assign these values are the size of the household and the age of its members (whether they are adults or children).’ But the equivalence scale is on shaky grounds theoretically when it comes to renters, as the economies of scale of consumption do not apply, pushing up the measured income available to these groups artificially. Rather than two sharing renters earning €15,000 a year each, their equivalised incomes become €18,000 a year each. Sharing renters tend not to share expenses and manage their finances individually and separately. They cannot draw on each other’s income and their housemate’s income has no bearing on their own living standards. The same even applies to most cohabiting couples. Data on the incomes of sharing renters would be more appropriately collected on an individual basis, reflecting actual access to income.
The earnings of a full-time minimum wage worker and most low-wage workers is not sufficient to rent alone throughout most of the country, clearly. A full-time minimum wage worker earns €1,575 a month (€18,900 a year) before tax and an average one-bed in West Dublin is €1,400 (in 2018, about 29 percent of Irish workers earned less than this) and €760 in Waterford City. For context, an Irish person is said to be at risk of poverty if they earn €973 or less a month (€11,676 a year) i.e. they would need a 50 percent wage increase to cover the rent (no bills or food) for a one-bed in Lucan or would have 50 euro a week to live on after rent in Waterford. Wage inadequacy therefore is increasingly forcing huge swathes of low wage earners to share rental accommodation or stay at home. A single low-wage worker (who hasn’t inherited property) is unlikely to live alone in a household in the bottom 10 percent of the ‘household’ income distribution for the very simple reason that their wages alone will not cover rent and bills in most parts of the country. The cut-off threshold for the bottom 10 percent was €13,600 and €16,600 for the bottom 20 percent in 2018. That’s right, the entirety of the income of the person at the 20 percent mark of the Irish income distribution (1 in 5 Irish people are in receipt of less income) would not cover rent for a one-bedroom apartment in Lucan. That is 31.6 hours a week at the minimum wage.
From a purely compositional perspective, this pushes up the ‘household’ income of any property shared by renters (and adult children living at home) but doesn’t really capture the relative standard of living (or the nature of their individual income situation) of those large sections of Irish society. As such, low wage workers tend then to end up in middle to upper middle income ‘households’ in the income distribution data, sharing with others. For example, the disposable income of a full-time minimum wage worker is €17,991. If this individual could afford to rent alone (maybe in Mayo or Donegal) this would place him/her in the 3rd decile (between the 20 and 30 percent marks) of the Irish income distribution. Meanwhile, the equivalised disposable income of two full-time minimum wage workers sharing is €22,488, which is in the fifth decile (middle-income by definition). The bottom few deciles are made up, for the most part, of pensioners (who are much more likely to own their own home and don’t have to worry about wage inadequacy relative to rent) and those not in employment. An increase in the minimum wage will therefore tend to benefit ‘middle to high’ income households.
This has particular implications for time-series analysis of income in an Irish context over the past decade or so. If three young, single professionals living together earned €15,000 each (after tax), an analysis of household income would show them as one household with an annual disposable income of €45,000 (equivalised to €19,400 each)2. An analysis of individual single households would classify the same individuals as three units earning €15,000 each, with obvious implications for any research on income distribution. This is a difference of three units in the second decile of the income distribution (equivalised) and one in the fourth decile.
Consider a young retail worker living alone on €16,000 (bottom 20% of household income distribution) had her hours cut so that she was coming out with €15,000 as the cost of rent for her 1 bedroom apartment (in rural Leitrim) went up 4 percent. If she moved home to her parents with a household income of €45,000 or €27,000 each equivalised (a middle-income household), income analysis by dwelling would report that two households at the bottom and in the middle became one in the middle. The equivalised disposable income of the new arrangement is €25,862 each (that’s an average of €3,862 more each and the young retail worker is now considered to have over 50% more income). Irish average equivalised disposable household income would actually rise due to this change in household composition. The income for individuals in this increasingly common situation and others like it (like when a student can’t move out in the first place) is obscured in household surveys (SILC in this case), which along with the rise in sharing renters are overstating the living standards of a huge chunk of low-income individuals. Wage inadequacy is artificially pushing up the incomes of these relatively low-income households.
This mismatch is also evident in Eurostat SILC data on household type, income quintile (placement in the distribution by fifths) and material deprivation rates3. Material deprivation is a key consumption based indicator to complement pure income measures to assess real material living standards. ‘3+ adult households’ have the second highest mean household incomes of any household type (others include single adult, 2 adults with children etc.) and the bulk of the category will be made up of individuals in the living situations discussed (though obviously many 2 adult households will be sharing renters or even single parent households with an adult child). Generally, income and deprivation have a strong negative relationship (income goes up, deprivation goes down). The deprivation rate for all households in the bottom 20 percent of the distribution is 30.4 percent (almost three times the average rate of 11.9 percent). This makes sense. With 3+ adult households however, this relationship breaks down. The mean equivalised disposable income of 3+ adult households was €34,053 in 2018, over €5,000 or 19 percent more per person than the average equivalised disposable income (€28,626). The material deprivation for 3+ adult households with no children in the third income quintile (the middle 20 percent) is 11.2 percent (over twice the figure for all third quintile/middle income households at 5.2 percent). These are by definition, middle-income, but the incidence of deprivation is almost as high as the national rate of 11.9%.4 Deprivation for 3+ adult households with no children in the 4th income quintile (getting in to ‘high-income’ here) is over three times the average for all households in that quintile and even higher than the deprivation rate of all ‘middle income households’ (3rd quintile). The share of 3+ adult households who would be unable to meet an unexpected expense is also not far from the average (32.4 v 37.3 percent) even with equivalised incomes so much higher than average.
It will be interesting to see whether the dominant narratives on Irish income inequality over the past number of years hold up once these changes are brought in. We’ll have to see. They will be especially important to get an accurate picture of changes to income distribution with the current upheaval.
Jeff Bezos added €13 billion to his net worth in one day in June (the entirety of the Irish health budget is around €17 billion) in an environment where 10’s of millions of Americans have lost their jobs. It’s clear there will be major disruption.
1 The CSO use a national scale which attributes a weight of 1 to the first adult, 0.66 to each subsequent adult (aged 14+ living in the household) and 0.33 to each child aged less than 14. The weights for each household are then summed to calculate the equivalised household size. The OECD assigns a value of 1 to the household head, of 0.5 to each additional adult member and of 0.3 to each child.
2 The OECD provide a note on the indicator: ‘The needs of a household grow with each additional member but – due to economies of scale in consumption– not in a proportional way. Needs for housing space, electricity, etc. will not be three times as high for a household with three members than for a single person. With the help of equivalence scales each household type in the population is assigned a value in proportion to its needs.
3 A household experiences material deprivation if they cannot afford 3 out of a list of 9 everyday expenses (Eurostat).
4 Eurostat define deprivation differently to the CSO. A household must be unable to afford 3 out 9 different everyday expenses (rather than 2 out of 11), hence the lower rate.