Jeremy Moore repairs long-haul trucks at a service centre on the interstate highway north of the Ohio town where he lives. Rather than being paid a fixed wage, he receives a commission for each truck he fixes. The fact that he and his wife Becky always welcome very hot weather or very cold weather has nothing to do with swimming or skiing: very hot bitumen burns out truck tyres, and very cold weather weakens batteries and alternators. During mild weather Jeremy’s take-home pay can drop by nearly half, from $3400 in a typical July to $1800 in March.
Janice Evans works the night shift in a casino on a reservation in Mississippi. She is guaranteed US$8.35 an hour, but in a good week she might receive twice that much including tips. When the weather is hot, people come to the casino for the air conditioning, but on many weekends in autumn they stay at home on Friday, Saturday and Sunday nights to watch football. Janice’s fortnightly take-home pay in 2012–13 varied between $1200 in July and $900 in February, and from September to November she had to cut back on food purchases.
Elsewhere in Ohio, Sarah and Sam Johnson have apparently stable middle-class incomes of roughly US$65,000 combined before taxes, but spend about US$6000 a year more than they earn. Their expenses — mortgage, insurance and food — are mostly predictable. But even though they have health insurance coverage from their employers, it is low-premium, high-deductibles coverage, meaning that they have paid significant out-of-pocket expenses for minor surgery, MRIs and an emergency room visit, and owe money on a specialised credit card for medical expenses. These expenses have been exacerbated by other unexpected costs: a major leak in a water pipe in their home, and one of their cars breaking down and having to be replaced by a cheap car that then required substantial repairs later in the year. While the Johnsons’ average monthly spending is around $4800, it was nearly $10,000 in one month and close to $8,000 in another.
During 2012 and 2013 a team of ten researchers lived in Ohio, Kentucky, California, Mississippi and New York, where they repeatedly interviewed these three families and members of another 232 households. Their aim was to track every dollar the householders earned, spent, saved and borrowed, every gift they gave or received, every donation they made to charity or friends, and any government benefits they were paid. The results were then analysed by Jonathan Morduch and Rachel Schneider for their book The Financial Diaries: How American Families Cope in a World of Uncertainty.
Morduch and Schneider’s focus is on income volatility — the extent to which a household’s income in any month can vary from its average monthly income over the year. They found that the average household in their sample experienced upward spikes in income during 2.2 months of the year and dips during 2.4 months, adding up to some degree of volatility during a total of 4.6 months. But a high-volatility group of households experienced spikes and dips during an average of 6.6 months a year.
Poorer households — those on around 42 per cent of median income in 2012 — tended to experience more and larger spikes and dips, ranging from 49 per cent below the monthly average to 58 per cent above. But even those classified as having “moderate incomes” (at least 84 per cent of median income) experienced fluctuations between minus 44 per cent and plus 49 per cent. Only 2 per cent of households recorded no significant monthly variation.
It wasn’t only incomes that ebbed and flowed. Spending was also volatile, partly because many families were effectively living from payday to payday, putting off paying bills in some months and trying to catch up in others. Many expenses — rent or a mortgage, for example — were predictable, but others — car or housing repairs and medical expenses — were unexpected.
As Sarah and Sam Johnson’s experience showed, healthcare costs are a big factor. According to Morduch and Schneider, the Johnsons were “unlucky but not so unusual”: about a quarter of the households had sent a family member to an emergency room during the year and about one in ten had a family member hospitalised.
The Financial Diaries makes up for its relatively small sample of households with its richly detailed accounts of how individual households cope with volatility. A broader picture comes from a large set of US data used in a JPMorgan Chase study of the income and consumption habits of 2.5 million bank account holders, transaction by transaction, between October 2012 and December 2014.
The study found that income and consumption were more volatile on a monthly basis than they were from year to year. Seventy per cent of the sample experienced annual income changes of more than 5 per cent between 2013 and 2014, but 89 per cent experienced monthly income changes of more than 5 per cent. Similarly, 41 per cent experienced income fluctuations of more than 30 per cent month-to-month but only 26 per cent experienced more than a 30 per cent change in annual income. Wages and salaries were the major contributor to volatility.
A study of further data from this source found that most of the month-to-month volatility in take-home pay occurred among individuals who stayed in the same jobs. While four-in-ten individuals experienced a job transition in a given year, this accounted for only 14 per cent of the month-to-month volatility in labour income.
The risk factors uncovered in the Financial Diaries — working on commission, relying on tips, coping with healthcare expenses in the absence of universal health insurance — may seem specific to the United States. But evidence from Great Britain reveals similar experiences there, albeit for different reasons.
In Tracking Income: How Working Families’ Incomes Vary through the Year, researchers at the London School of Economics enlisted 180 householders to report on their income and expenditure during the 2003–04 financial year. Nearly half dropped out of the study during the year, but complete week-by-week information was available for the remaining ninety-three.
Of those families, only seven had a stable weekly income over the course of the year, with another twenty-one having “broadly stable” income (within 15 per cent of the annual average for ten or more four-week periods out of the full thirteen periods). Thirty-two families stayed within 15 per cent of the average for ten or more periods but had large “blips” in other periods. Twenty-six families had erratic income patterns — eight of them highly erratic, with four-week earnings varying by up to £1000 per month, or in one case from a high of £2500 to a low of zero.
These variations reflect changes both within the household (in partnership status or number of children) and at work (starting, changing or stopping jobs). While benefits and tax credits generally helped stabilise incomes — rising when earnings fell and vice versa — in some cases they amplified income changes.
A more recent report on irregular pay from the Resolution Foundation, a British think tank, used anonymised transaction data from over seven million Lloyds Banking Group accounts. It found that monthly pay fluctuations are the norm for the majority of employees, with only the 9 per cent of employees who remained with the same employer throughout 2016–17 experiencing monthly fluctuations in take-home pay of more than 5 per cent. More than four in every five steadily employed low-income earners on around £10,000 per year had volatile pay, compared with two-thirds of those on around £35,000 a year. The absolute average monthly pay change for those with a steady job was highest (in excess of 15 per cent) for those on the very lowest earnings.
Do the same patterns exist in Australia? The short answer is that we don’t know — partly because of a lack of data.
The Household, Income and Labour Dynamics in Australia survey, or HILDA, provides very detailed and useful information on income changes over time — but only between years. Using HILDA data for 2002–07, the Income Inequality, Mobility and Economic Insecurity in Australia study found that the lowest 10 per cent of earners carried over 30 per cent of total volatility risk and the lowest 20 per cent accounted for 46 per cent of lost wellbeing due to income uncertainty. It seems highly likely that changes within years would add to overall income volatility, as the JPMorgan Chase study found.
Drawing on the Australian Bureau of Statistics’s 2016 Characteristics of Employment Survey, researchers Iain Campbell and John Burgess found that more than half of all casual workers in Australia and 15 per cent of permanent employees reported earnings (excluding overtime) that varied from one pay packet to the next in their main job. The ABS’s 2018 survey found that 24 per cent of all employees reported earnings varying from one period to the next (excluding overtime payments), 21 per cent often working a different number of hours each week, and 19 per cent with no guaranteed minimum number of hours each week. The survey doesn’t tell us, however, how much their earnings varied from week to week.
A recent large-scale study surveyed 1101 Australian adults who had been unable to pay a debt when it fell due within the previous two years. Of the participants, around 44 per cent were employed, 36 per cent derived their income primarily from social security payments, and around 7 per cent received both wages and a Centrelink payment. The most common tipping points into financial hardship included unforeseen expenses (37 per cent), a reliance on income support (33 per cent), unexpectedly large utility bills (27 per cent), physical health problems (27 per cent), mental health problems (22 per cent), not enough work (20 per cent) or unemployment (more than six months, 17 per cent; less than six months, 10 per cent). Significantly less prevalent were factors such as gambling (4 per cent) and alcohol and drug addiction (3 per cent).
While this study suggests that income volatility could be a factor for those without enough work, it doesn’t measure income changes directly. One study that did try to do that was a survey of seventy low- and moderate-income households in suburban Melbourne by the Brotherhood of St Laurence and RMIT University. It found that over half of the participants experienced highly erratic variations (greater than 25 per cent) in their fortnightly incomes. One group of nine households had variations greater than 60 per cent.
An analysis of those findings noted that many of these households reported having provided incorrect wage estimates to Centrelink. Casual work and fluctuating hours of work make it hard for households to estimate incomes accurately, and Australia’s highly income-tested benefit system can magnify the impact of reporting errors by cutting benefits in the immediate fortnight (if the estimate of earnings was too high) or in the subsequent fortnight (if the estimate was too low). Most of the study’s respondents sought to cope by reducing spending on food, recreation, utilities, medical care and transport. Just over a third borrowed money from family or friends.
But a comparison of two other studies suggests that Australian households are less likely to experience financial stress than American households. Since 2002, the US Federal Reserve Board’s Survey of Household Economics and Decisionmaking, or SHED, has been providing a wide-ranging overview of the economic wellbeing of American households, including how they deal with unexpected expenses. SHED found that 27 per cent of American adults were forced to skip some form of medical treatment in 2017 (compared with 32 per cent in 2013). Skipping dental care was most common, with 19 per cent going without treatment because of cost in 2017.
Australia’s HILDA survey has collected information on indicators of financial stress in Australia every second year since 2001. The results for 2014 show that about 1.1 per cent of its sample did not have medical treatment when needed, 0.5 per cent went without medicines prescribed by a doctor, 5.2 per cent went without dental treatment and 3.3 per cent went without a yearly dental check-up for their child. These deprivations could potentially overlap, but even cumulatively they show many fewer difficulties with medical expenses than in the United States — not a surprising result given the difference in healthcare systems.
While the measures may not be fully comparable, Australian data also show that emergency expenses generally created fewer difficulties for households. In 2015–16 around 13 per cent of Australian households said they would be unable to raise $2000 in a week for something important. In the United States, the 2017 SHED found that 41 per cent of American adults would either borrow or sell something or not be able to pay if faced with a $400 emergency expense (a reduction from around 50 per cent in 2013). Of this group, 29 per cent — 11.9 per cent of all households — said that they would not be able to pay the expense right now. The percentage of households is slightly lower in the United States, but the amount of money is much lower.
Inequality of wealth is a significant factor. While wealth is much more unequally distributed than income in all countries, the disparities are much greater in the United States than in Australia. According to the latest Credit Suisse Global Wealth Report, mean wealth per adult is about 20 per cent higher in Australia than in the United States, but median wealth in Australia, at $191,000, is nearly four times higher. About 6 per cent of adult Australians have less than $10,000 in wealth; in the United States, the corresponding figure is 28 per cent. Put another way, the least wealthy 60 per cent of Australians have 17.2 per cent of total wealth whereas the least wealthy 60 per cent of Americans have 3.2 per cent of total wealth.
Yet a high share of Australians’ wealth is held in the form of housing and superannuation, neither of which is readily available to meet unexpected expenses. In 2015–16, the least wealthy 20 per cent of Australian households had only $10,000 in financial assets apart from superannuation, and owed $13,000 on student loans, credit cards and/or motor vehicle loans (not including mortgages), suggesting considerable vulnerability for this group.
American wealth data rank households by their incomes rather than by their net worth. On this measure, the poorest 20 per cent of US households had an average of US$6700 in net financial assets. In Australia, the figure was A$63,200. (The difference between the rankings by income and wealth reflects the fact that low-wealth householders tend to be young people who have not had time to accumulate financial assets or housing, while the lowest income group is made up of older people who have had many years to accumulate savings and pay off their debts.)
The very modest wealth held by the bottom half of American households is likely to both reflect and reinforce the greater likelihood of financial stress, which in turn will be influenced by the different levels of public coverage of health costs, the level of the minimum wage, and the conditions of employment faced by workers.
The Financial Diaries provides important insights into the challenges of income volatility for American households. It also clearly has implications for a wide range of other public policy challenges — for financial inclusion, for minimising recourse to payday lenders, for relieving household indebtedness, for providing financial counselling, and for increasing the effectiveness of insurance. Volatile incomes pose challenges for income-tested benefit programs and for the collection of taxes to finance social programs. Volatility may also create a climate in which wage demands fall, wage growth stalls — and budget projections are less accurate.
The labour markets of Australia and the United States are different in important ways, as are our social and economic institutions, and the Australian welfare system seems to be more effective in offsetting the risks associated with volatile incomes. Australia’s more comprehensive healthcare coverage seems particularly important. Australia’s tighter employment protections and higher minimum wages could be expected to lessen the impact of income volatility and may be associated with the very different distributions of wealth in Australia and the United States, and thus the different availability of assets to cushion against income shocks.
But Australian households do face significant income risks. HILDA’s longitudinal data show that illness and disability are particularly common, with around 40 per cent of the Australian population experiencing a serious personal injury or illness each year over a ten-year period, and nearly 70 per cent of men and 64 per cent of women needing to cope with a close relative or family member’s serious injury or illness in any ten-year period. Over a nine-year period in another HILDA report, 22 per cent of men and 16 per cent of women were dismissed from their job. Thirty per cent of people who were twenty-five or younger at the survey’s start were dismissed from their jobs over the period. As a result, only 2.2 per cent of the Australian population stayed within the same percentile of the income distribution between 2001 and 2010. According to the latest HILDA survey, around a quarter of working-age men and nearly a third of working-age women experienced a period of income poverty in the ten-year period after 2001.
We also know that Australia has one of the highest levels of part-time employment of any high-income country and the highest level of underemployment in the OECD. A 2015 OECD report, In It Together, found that the share of part-time temporary employees in Australia is the highest of all the countries included, being four times the OECD average and nearly twice the next-ranked countries, the Netherlands and Japan.
The absence of detailed figures on month-to-month income volatility — which we know is likely to be more common than year-to-year volatility — is a glaring lack in Australian social data. It’s one that will need to be remedied if we are to avoid the worst consequences of precarious labour markets. •