Chapter 4. The rising cost of the middle-class lifestyle

This chapter examines how the living standards of the middle class have evolved over recent decades. It also assesses the contribution of middle-income households to aggregate consumption in the economy, in comparison to its shares of total population and income and those of the lower- and upper- income classes. The chapter then examines to which items middle-income households allocate their budgets. It goes on to consider how middle-income household spending has increased relative to income, what consumption categories have driven that increase, and to what extent prices have been an influencing factor. Finally, it considers the ability of middle-income households to afford their standard of living, their exposure to financial vulnerability and their over-indebtedness.

The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.

Introduction and key findings

A middle-class lifestyle is typically associated with certain goods, services and living conditions, such as decent housing, good education, and sound health and a wholesome environment (Box 4.1). However, the prices of some of those goods and services have risen much faster than general inflation. In the last two decades, higher education and house prices in particular have outpaced CPI inflation and household median income.

This chapter gathers cross-country evidence on the extent to which the living standards of middle-income households1 have been squeezed in recent decades. Section 4.1 assesses the contribution of middle-income households to aggregate consumption in the economy, relative to its share of total population and income, and in comparison to the lower- and upper-income classes. Section 4.2 examines how middle-income households allocate their budget across categories of consumption. Section 4.3 considers how middle-income households’ spending has increased relative to their incomes, what consumption categories have driven that increase and to what extent prices have been a factor. Section 4.4 considers the ability of middle-income households to afford their standard of living, their exposure to financial vulnerability and over-indebtedness. The final section presents concluding remarks.

From the analysis and evidence in this chapter, the following findings emerge:

Middle-income households contribute more than any other income group to aggregate consumption in the economy. With almost two-thirds of total consumption, middle-income households are by far the income group that accounts for the highest share of consumption. Shares range from 50% or below in the United States, Southern European and Latin American countries to 70% or above in the Nordic countries and Slovakia. Middle incomes are the only class whose share of expenditure is larger than both its shares of income and its share of population.
Middle-income households’ expenditure increased faster than their incomes prior to the global financial crisis, plummeted between 2007 and 2010, and has remained stagnant since then. Like other recessions, the Great Recession saw a steep decline in the middle-income spend on durables. Unlike previous recessions, however, there was also a fall in middle-income spending on non-durables, especially services.
Core goods and services make up the bulk of middle-class spending. Middle-income households spend more than half of their budget on core items such as housing, food, clothing, health and education. Leisure spending (including recreation, culture, hotels, restaurants and alcohol) accounts, on average, for 12% of middle-class budget.
Almost one-third of the budget of the average middle-income household goes to housing, on average across OECD countries. Imputed rents on owned homes are their single largest expenditure, as most middle-income households are homeowners. Food and clothing account for slightly less than one-quarter of the budget of the average middle-income household in OECD countries.
Housing has been the main driver of rising middle-class expenditure. Its share of middle-income budgets increased OECD-wide from one-quarter to almost one-third between 1995 and 2015. Furthermore, house prices have grown twice as fast as inflation and one-and-a-half times faster than the household median income.
Middle-class households assign an increasing share of their budget to private health-related expenses as they spend more on out-of-pocket and private health insurance.
In some countries, the rising costs of higher education threaten the ability of the middle-income class to send their children to university, as they have increased faster than inflation and median incomes.
In addition to rising prices, changes in consumption preferences can explain some of the shifts in middle-income budget expenditures. In fact, some middle-income households seek to match the spending patterns of high-income households. So, as the rich get richer, the middle classes spend more and more.
As living costs have increased and expenses grown faster than incomes, many middle-income households find themselves struggling to make ends meet, grappling with financial vulnerability, and even dissaving (i.e. spending more than they earn). Furthermore, a considerable share of middle-income households have excessive debt – one in eight hold debt that is greater than three-quarters of their assets.

Box 4.1. The standard-of-living approach

Most studies into the middle class use income-related indicators. This is in line with the so-called “resource-based” concept of economic well-being, according to which income is a fundamental determinant of a household’s economic situation. An alternative concept is “standard-of-living”. It is based on consumption-related indicators, which claims that a household’s use of resources is more representative of its ability to satisfy its needs than the resources themselves.

At the core of the distinction between these two approaches are two important notions of economic theory – the permanent income hypothesis (Friedman, 1957), and the life-cycle hypothesis (Modigliani and Brumberg, 1954). According to the permanent income hypothesis, individuals attempt to keep consumption relatively smooth over their lifetime, basing their consumption decisions on what they consider their permanent long-term income will be rather than on their current income. In the same vein, the life-cycle hypothesis suggests that individuals plan their consumption and savings over their life cycle, smoothing out fluctuations in income in each phase of the cycle. According to these hypotheses, any income changes perceived as temporary will not induce significant consumption changes. Measures based on current income thus yield an assessment of transitory changes to economic well-being, while measures based on consumption will be more representative of long-term economic well-being.

Despite the strong conceptual case for a joint analysis of income and consumption, there are very few examples in the literature of an analysis of this type. This is related largely to data and methodological issues regarding the definitions and uses of consumption measures (Aguiar and Bils, 2015; Attanasio and Pistaferri, 2016) and to the fact that detailed information on both income and consumption is seldom available in the same data source. Income data are easily accessible and comparable over time. Moreover, international standards for the definition and measurement of income variables have long been established, making it reasonably easy to perform comparisons across countries. This is less the case for consumption data. Nevertheless, considerable efforts have been made recently to improve the availability of consumption data particularly through the introduction of common international guidelines for consumption surveys (OECD, 2013[1]) and the use of new data sources and methods (Browning, Crossley and Winter, 2014).

In accordance with the recommendations by the 2009 Stiglitz-Sen-Fitoussi Commission (Stiglitz, Sen and Fitoussi, 2009[2])and the Vienna Memorandum signed by the Directors Generals of National Statistical Institutes, in 2017, the OECD and Eurostat launched a joint Expert Group (EG ICW) to develop experimental distributional measures of income, consumption and wealth at the micro-level. These data may help describe people’s economic well-being more thoroughly and improved understanding of the dynamics of inequalities (OECD, 2018[3]).

4.1. Middle-income households contribute the most to consumption

Middle-income households contribute the most to the aggregate consumption in the economy. With almost two-thirds of total consumption, middle-income households are the income group that accounts for by far the highest share of consumption, ranging from 50% or below in the United States, Southern European and Latin American countries to 70% or above in Nordic countries and Slovakia. Besides its sheer size, the middle-income class has, in comparison to other income classes, both the resources and the propensity to consume. Middle incomes are the only income class whose share of spending is greater than both its share of income and its share of the population (Figure 4.1, Panel A). Lower-income households spend proportionally more than their share of total income, but less than their share of the population. Higher-income households spend proportionally more than their share of the population but less than their share of total income. In Chile and Mexico, as well as Brazil and South Africa, middle-income households account for less than half of overall consumption expenditure and spend proportionately less than their shares of the population. In Brazil and South Africa, upper-income households contribute more than middle incomes to overall consumption spending (Figure 4.1, Panel B).

Figure 4.1. Middle-income households account for the largest share of overall consumption expenditure

Box 4.2. Household consumption expenditure: measurement and data

The analysis in this chapter draws on micro-data relating to household consumption expenditures. The data are sourced from household budget surveys. The expenditure categories available in national surveys have been harmonised by the OECD in accordance with the Classification of Individual Consumption according to Purpose (COICOP) (United Nations, 2018[4]). This classification, developed by the United Nations Statistics Division, divides consumption into categories, with twelve main categories for the household sector – Annex Table 4.A.1 describes the COICOP main categories and subcategories used. Data used follows the COICOP classification, with the exception of “insurance related to health”, classified as part of “health” consumption instead of “miscellaneous goods and services”.

For simplicity, some categories are referred to together with others. So, the sum of “Recreation and culture”, “restaurants and hotels” and “alcoholic beverages, tobacco and narcotics” are referred to as “leisure”. The sum of “food and non-alcoholic beverages” and “clothing and footwear” are referred to as “food and clothing”. The sum of “furnishings, household equipment and routine household maintenance” and “Miscellaneous goods and services” (excluding “insurance related to health”) are referred to as “other”.

Data were collected for 28 OECD countries (data are not available for Australia, Canada, Iceland, Israel, Japan, Korea, New Zealand and Switzerland), and two emerging economies (Brazil and South Africa) – see Annex Table 4.A.2 for an overview of data sources. Data for EU countries, South Africa, Chile and Brazil were already compatible with COICOP. Data for the United States and Mexico were reclassified in accordance with COICOP. Data for EU countries for 1998, 2005 and 2015 were imputed with growth rates by quintile of disposable income based on data published by Eurostat (except for France, the United Kingdom and Spain between 2006-2015). Incomes have been adjusted for non-reporting using experimental statistics on income published by Eurostat (Eurostat, 2019[5]).

An important issue to factor into analysis of consumption patterns is the distinction between actual consumption and consumption expenditure. The latter is typically a subset of the former, and yields only consumption data that have been obtained through the purchase of goods and services in markets. There are several aspects to the difference between actual consumption and consumption expenditure.

Home production of services for own use may account for a non-negligible part of a household’s consumption. It is reflected in consumption expenditure only through the value of the purchase of inputs used for home production, which is less than the market value of the services.
The provision of free or subsidised goods and services provided to households by governments and non-profit institutions is not registered as consumption expenditure in household budget surveys. For example, in the event of the free provision of public education or medical care services, household expenditure on education and health is lower than the actual consumption thereof.
Spending on consumer durables leads to differences between consumption and consumption expenditure: while expenditure on durables is sporadic, consumption thereof is spread over multiple periods. This difference is typically accounted for by converting expenditure into the rental value of these types of goods over time. In COICOP, this approach is applied to the purchase of a home. Instead of the amount of the purchase or mortgage repayment, COICOP measures the consumption of an owned house by imputing a rent (United Nations, 2018[4]). The imputed rents of owned houses are available for all countries and years, except for Estonia and the United Kingdom in 2010 and 2015.

4.2. The middle class spends most of its budget on core goods and services

Middle-income households spend more than half of their budget on core items such as housing, food, clothing, health and education. Housing is the largest middle-class spending item in most OECD countries, followed by food and clothing, transport, leisure, others, health, communication and education (Figure 4.2, Panel A). As expected, lower-income households devote a higher proportion of their expenditure to core needs, while upper-income households are able to devote a higher share to other items such as transport and leisure.

An average of almost one-third of the budget of middle-income households goes on housing in OECD countries. In Hungary, the share is almost 40%, while in Latvia and Mexico it is less than 25% (Figure 4.2, Panel B). The imputed rents from owned homes are the largest single item of housing spending, as most middle-income people are homeowners.2

Payments on mortgages (interests plus principal repayment) account, on average, for almost 20% of the disposable income of middle-income households with mortgages. From country to country, mortgage-to-income ratios of middle-income households range from less than 10% in the Czech Republic and Sweden to almost 30% in France (OECD, 2019[6]). Such differences reflect variations in outright home ownership, access to mortgages, mortgage conditions, loan-to-value and loan-to-income ratios (André, 2016[7]; Whitehead and Williams, 2017[8]).

Similarly, actual rents comprise, on average, almost 20% of the disposable income of middle-income tenant households renting in the private sector (OECD, 2019[6]). This burden ranges from 4% in Latvia3 to more than one-quarter in Chile, the Czech Republic and Greece. On average, tenants in the private sector make up about 20% of middle-income households, ranging from less than 10% in Estonia, Hungary, Latvia, Lithuania, Norway, Poland, Slovak Republic, Slovenia and Spain to more than 40% in Denmark, Germany, Korea and Switzerland.

Food and clothing account for slightly less than one-quarter of the budget of middle-income households, on average across OECD countries. In the Czech Republic, Latvia, Lithuania, Mexico and Turkey, as well as in Brazil and South Africa, food and clothing make up almost one-third of the budget of middle-income households. Leisure accounts, on average, for 12% of household spending. However, the share is particularly high in Austria and Sweden, where leisure spending is almost as high as food and clothing. It is modest, by contrast, in Hungary, Lithuania, Mexico, Poland and Turkey, as well as in Brazil and South Africa. Transport accounts for almost one-fifth of middle-class spending in Norway, Slovenia and the United States. Only in Greece, Hungary, Lithuania and Poland, middle-income households spend less than a tenth of their budget on transport. Since health care and education are publicly provided in most countries, household spending on these items tends to be limited. However, in the United States health accounts for 9%, while in Mexico education accounts for 8% of the average middle-income household budget.

Figure 4.2. Core items account for the bulk of middle-class budgets

4.3. As the cost of core goods and services has risen, so has middle class spending

4.3.1. Middle-income spending has risen faster than income

Prior to the global financial crisis, the consumption expenditure of middle-income households increased faster than their incomes (Figure 4.3, Panel A). Middle-income expenditures plummeted between 2007 and 2010 and remained stagnant since. As in other recessions, the fall was steeper in spending on durables. Unlike previous recessions, however, middle-income spending on non-durables (especially services) also declined (De Nardi, French and Benson, 2012[9]; Crossley, Low and O’Dea, 2013[10]). The fall in asset prices and consumer confidence may explain a considerable part of the steep decline (Petev, Pistaferri and Eksten, 2011[11]).

In Greece, Ireland and Spain, where the crisis was particularly harsh, expenditure increased strongly between 1995 and 2005 and fell sharply between 2005 and 2015 (Figure 4.3, Panel B). In Australia, Belgium and the Netherlands, by contrast, middle-class income and spending grew in both periods. In Austria and Portugal, it also grew in both periods, despite a considerable fall between 2010 and 2015.

In the United States, median expenditure grew faster than median incomes during the pre-crisis period. Between 2008 and 2010, median spending decreased in line with the fall in median incomes. From 2011, as the economic recovery began, median expenditure quickly returned to pre-crisis levels, while incomes lagged behind. In 2004, households in the middle fifth of the income distribution spent 90% of their income. By 2016, the share had climbed to nearly 100%, which suggests that American middle-income households are no longer saving (PEW, 2016[12]).

In Australia, household spending increased much faster than inflation in the last 30 years. Between 1984 and 2015-16 average weekly household expenditure on goods and services almost tripled, while the consumer price index (CPI) almost doubled (ABS, 2017[13]).

In the United Kingdom, expenditure declined more among young than old people and more among mortgage holders than tenants. Policies implemented during the period to stimulate consumption, such as the temporary reduction of VAT and a subsidy to purchase new cars, seem to have brought forward spending that would have happened later. In fact, a sharp increase in the consumption of durable goods was observed during the period these measures were in place (Crossley, Low and O’Dea, 2013[10]). In Spain, between 2007 and 2015, consumption fell more steeply among households with high than low consumption. A possible explanation could be that worse-off households made better use of economies of scale than the better-off: while the size of the average household fell in both groups, it fell considerably more among the better-off (Cardoso and Montañez, 2017[14]).

Figure 4.3. Middle-income class spending has increased faster than income

4.3.2. The middle-class is spending more on core goods and services

The rise in middle-income spending was driven mainly by core goods and services. Housing and, to a lesser extent, health were the main items to increase their share on the budget of middle-income households in the last two decades (Figure 4.4). The budget shares of transport, communications and education remained fairly stable, while those of food, clothing, leisure and other items fell considerably.

Figure 4.4. Middle-income class spending on housing and health has increased

Housing makes up an increasing share of middle-class budgets

Besides being the largest spending category (Section 4.2), housing has been the main driver of rising middle-class expenditure in recent decades. Across OECD countries, the share of spending that middle-income households devote to housing increased from a quarter in 1995 to almost one-third in 2015 (Figure 4.5, Panel A).

Propelled by a strong rise in house prices (See section 4.3.3), owned homes contributed the most to the rise in housing spending, particularly in the period 1995-2000 and 2010-2015. Consumption on owned homes rose particularly steeply in Ireland, Portugal and Spain, while housing expenditures has fallen in Belgium, Chile and Finland (Figure 4.5, Panel B).

Although to a less extent than owned homes, spending on rents and utilities also contributed to housing’s higher share of middle-income budgets, particularly between 2005 and 2015. In 2015, spending on rents accounted for an average of 6% of middle-income household expenses, 2 percentage points more than in 1995. The share of rents in household budgets was particularly high in Finland, Germany and Ireland, while it fell in Greece and the Netherlands (in comparison to two decades earlier).

Utility services have also contributed to higher middle-class housing expenditure. On average, utilities such as water, fuel and electricity accounted for 7% of middle-income household budgets in 2015, 1.5 percentage points more than in 1995. Utilities make up particularly large shares of middle-class spending in Eastern European countries. The large stock of energy-inefficient buildings may explain why they are so large in comparison to countries with similar climates such as Finland and Sweden (OECD, 2019[15]; 2018[16]; 2011[17]).

Figure 4.5. Housing has increased as a share of middle-income class budgets

Private health spending accounts for a growing share of middle-class budgets

Middle-class households devote increasingly greater shares of their budgets to health-related expenses – i.e. private health insurance and out-of-pocket expenditure on health products and services. In 20 OECD countries, middle-income households have increased their health-related spend by an average 0.5 percentage points since 2005, and by 2.5 percentage points since 1995 in 11 countries (Figure 4.4).

Between 2005 and 2015, the rise was particularly high (2 percentage points) in Chile, Germany, Latvia, the Slovak Republic, Spain and the United States. In the mid-1980s, health-related spending comprised a small fraction of the budget of middle-income households in the United States. By 2014 it had increased by 60%, in real terms, and is smaller only than housing, food and transport (Schanzenbach et al., 2016[18]). In Spain, more people (1 million between 2012 and 2017, a 10% rise) have enrolled in private health insurance following the cuts in the public health system introduced as part of fiscal consolidation measures (Alfageme, 2019[19]). In contrast, the share of middle-income household budgets devoted to health-related expenses fell 2 percentage points in the Netherlands and 1 point in Hungary.

Middle-class households are spending more on private health insurance than a decade ago. In the 18 OECD countries with available data, middle-income budget shares spent on health insurance has increased in all but four countries. In many countries, however, the level and rise of private health insurance spending are limited. Only in the Netherlands4 and the United States does private health insurance account for more than a 5% share of middle-income household budgets. Only in the Slovak Republic and the United States has that share increased by more than 1 percentage point. The population covered by private health insurance has increased in some OECD countries over the past decade, particularly in Denmark, Korea, Slovenia and Belgium, and has fallen in others, notably Greece, Ireland, New Zealand and the United States (OECD, 2017[20]).

The share of middle-class household budgets devoted to out-of-pocket health spending has increased in the last decade, on average across OECD countries. On a country level, it has fallen in around as many countries as it has risen, although rises tend to be stronger than falls. In Chile, the Czech Republic, Estonia, Latvia and Spain it increased by at least 1 percentage point between 2005 and 2015.

Figure 4.6. Middle-income households are spending more of their budgets on private health insurance and out-of-pocket health items and services

4.3.3. Prices of basic needs have soared

Achieving a middle class lifestyle has become more difficult than in the past because of the steep rise in house prices and other large items of middle-class consumption (Blank, 2010[21]). Together with changes in the amount and quality of goods consumed, the higher prices of such items drive spending shifts.

House prices outstrip inflation and median incomes

On average, in the last two decades, house prices have grown twice as fast as inflation and 50% more than the household median income, in OECD countries with available data (Figure 4.7, Panel A). After rapid growth in the early 2000s, the gap between house prices and both CPI inflation and household incomes narrowed during the global financial crisis but has widened again in recent years.5 By 2016, the latest year for which data are available, the differences between the growth rate of house prices and those of CPI inflation and household median income were the largest since records began.

In most countries, average house prices grew considerably faster than median incomes. In Denmark, France, New Zealand, Norway, Sweden and the United Kingdom, they increased at least 60% more than median incomes. In all those countries, house prices increased sharply between 1995 and 2007, plummeted during the global financial crisis and then recovered. The one exception was France, where house prices have remained somewhat stable since 2010, despite recently booming in Paris and other large French cities (Les Echos, 2018[22]). In Australia, Belgium, Canada, Israel, Norway, New Zealand and Sweden, real house prices in 2017 were higher than before the crisis.

Only in Germany, Ireland, Italy and Japan rises in house prices since 1995 not exceeded those of median incomes. In Germany, real house prices remained stable throughout the 2000s and are currently not much different from three decades ago, despite an acceleration in more recent years (Wijburg and Aalbers, 2017[23]). In Ireland and Italy real house prices experienced strong downward corrections in the wake of the global financial crisis. In Japan, after steep increases up to around 1990, house prices fell strongly between 1995 and 2010, since when they have slightly recovered (Engsted, Hviid and Pedersen, 2016[24]).

While real estate prices have been a major factor in the rise in spending on housing, changes in consumption patterns have also contributed. The increasing concentration of economic opportunities and better-paid jobs attract more people to large urban areas (Glaeser, 2010[25]; Combes and Gobillon, 2015[26]). Demand for good public services and gentrification have also nudged middle-class households into more expensive segments of the housing market (Chapple, 2016[27]; Hamilton, 2014[28])

Healthcare costs grow faster than inflation

Healthcare costs have, on average, increased faster than general inflation but less than median incomes in OECD countries. Between 1995 and 2005, they grew strongly at a similar pace to median income. After a slight decline during the global financial crisis, they have remained fairly stable. At the country level, though, there are variations. They increased considerably in real terms in the Czech Republic and Sweden, where real median incomes grew even faster. Taking into account differences in income growth between countries, rises in real healthcare costs were most significant in the Czech Republic, Japan and the United States. In contrast, they fell in France and Luxembourg, and increased less than median incomes in Australia, Israel and Norway.

Higher education costs increase student debt as university attendance grows

In some countries, rising costs prices for higher education are pushing more young people into debt. Between 1995 and 2015, the average cost of education-related goods and services has increased faster than CPI inflation and median incomes OECD-wide. The rise was driven primarily by higher education, as private spending in primary and secondary education is very low in most countries. Alongside the decline of middle-skilled jobs and the rise in numbers of highly skilled occupations, an increasing share of young people enter higher education, which spells substantial tuition fees in some countries.

Education costs increased most steeply in the United Kingdom where reforms in recent decades led to higher university tuition fees, as government funding for higher education shifted from up-front grants to universities to student loans (Dearden et al., 2017[29]). In the United States, the cost of education has also increased considerably faster than overall inflation and median incomes. By contrast, in the Czech Republic and Israel, the cost of education rose only very slightly more than inflation and less than the median income.

Other budgets items vary as spending patterns and preferences evolve

Trends in prices and spending have evolved differently across other categories. The decline in clothing’s share of budgets (Figure 4.4) may be associated with the fall in its relative price (Figure 4.7). By contrast, the fall in the share of the household budget devoted to food reflects lower real expenditures, while the relative prices of food have hardly changed.

Figure 4.7. House prices have increased much faster than inflation and median incomes

Changes in consumption preferences can explain some of the shifts in the composition of middle class budgets. In fact, recent studies have found that middle class spending patterns may be affected by changes in the income and expenditure of high-income households (see Box 4.3).

Box 4.3. Trickle-down consumption

Income inequality may increase middle-class spending

Wide income inequality may harm middle-class living standards by distorting consumption patterns (Frank, 2013[30]). Recent studies suggest that an increase in expenditure by an income group may prompt a rise in the expectations of higher living standards (and subsequently the spending patterns) among the income groups below. A study for the United States concluded that “rapid growth of income among top earners in recent decades has stimulated a cascade of additional expenditure by those with lower earnings” (Frank, Levine and Dijk, 2014[31]). Another study found evidence of “trickle-down consumption”, by which non-rich households increase their expenses on status-maintaining goods and services as top income levels rise. The study estimates that non-rich households would have saved up to 3% more annually if top incomes had grown at the same pace as the median (Bertrand and Morse, 2016[32]).

However, the consumption patterns of middle- and top-income households may be shifting apart. While the middle class increases spending on “conspicuous consumption” (e.g. apparel, watches, jewellery, cars and other socially visible goods), a new elite has moved its expenditure pattern towards “inconspicuous cultural markers”6 that convey the acquisition of knowledge rather than material goods (Currid-Halkett, 2017[33]). In the United States, between 1996 and 2014, the top 1% increased the share of income spent on education from 2% to 6%. Meanwhile, middle-income households still spend about 1% of their income on education, much the same as two decades ago (Currid-Halkett, 2017[33]).

Empirical evidence on middle class conspicuous consumption is, however, still far from conclusive. Analysis using data from France and the United States since the mid-1970s finds no evidence of a disassociation between social class and lifestyles. Instead, it suggests strong and persistent differences in spending patterns by social classes. Overspending would result from the necessity of meeting basic needs driven by increasing income inequality rather than from spending for positional and cultural distinction (Petev, 2013[34]).

4.4. Middle class increasingly financially vulnerable and at times over-indebted

As living costs rise and expenses increase faster than incomes, many middle-class households find themselves financially vulnerable and some with excessive levels of debt.

4.4.1. Many middle-income households are financially vulnerable or struggle to make ends meet…

Four in ten are financially vulnerable and half struggle to make ends meet

Almost 40% of middle-income households in 18 European OECD countries are financially vulnerable – i.e. they are in arrears or unable to cope with unexpected expenses or sudden falls in income. Shares vary widely from country to country, ranging from 12% in Norway to 70% in Greece (Figure 4.8). On average, the risk of middle-income households being financially vulnerable is closer to the risk run by the upper- than the lower-income class. However, in Greece and Hungary, the shares of middle-income households in financial vulnerability are much closer to shares among lower-income households.

Figure 4.8. Financial vulnerability affects four in ten middle-income households

In the 24 European OECD countries with available data, an average of 47% of middle-income households reported difficulty in making ends meet in 2016 (Figure 4.9). The figure was less than at the peak of the economic crisis (54% in 2012), but slightly higher than the pre-crisis period (46% in 2006). There is wide variation between countries, ranging from around 10% in Norway and Sweden to more than 70% in Hungary, Italy, Latvia and Lithuania, and 95% in Greece. While cross-country differences may reflect in part national traits and culture, it is significant that the incidence of middle-income households reporting trouble in making ends meet is closer to the incidence among the lower- than the upper-income class. This is particularly true of countries where large shares of middle-income households say they struggle. In Hungary, for example, the share is 78% among middle-income, 88% among lower-income and 36% among upper-income households. The incidence of struggling middle-income households is also closer, or relatively close, to the incidence of struggling lower-income groups in some countries where relatively low shares of middle-income households report difficulties. Examples are Denmark, Finland and Iceland, where shares have increased in comparison to ten years ago.

Figure 4.9. Half of middle-income households struggle to make ends meet

4.4.2. …and some have trouble maintaining their living standards

More than one in five middle-income households dissave, i.e. spend more than they earn, which carries the risk of over-indebtedness. In OECD countries, the share of middle-income dissaving households ranges from less than 10% in Estonia, Lithuania and Poland to more than half in Chile and Greece (Figure 4.10). However, the situation is not as bad as for the low-income class: in all OECD countries, the shares of middle-income people spending more than they earn is closer to those in the upper-income class. For example, in Sweden, 10% of middle-income and 1% of upper-income households dissave, in comparison to almost half of lower-income households. The share of middle-income dissavers has fallen since 2010, particularly in Eastern Europe. Some exceptions are Chile, the Slovak Republic and the United States, where shares have increased since the crisis despite the economic recovery.

Figure 4.10. More than one in five middle-income households spend more than they earn

4.4.3. Middle-income class more over-indebted than the total population

In many OECD countries, it appears that middle-class households used debt to finance consumption during the pre-crisis period. In the United States, for instance, debt levels in middle income quintiles increased dramatically relative to their net worth and income in the two decades before 2007. During the crisis, the middle class was also hit harder by the fall in house prices and higher mortgage debt relative to the value of homes. Following the crisis, middle-income households have drained their assets in order to repay debts and maintain consumption (Wolff, 2014[35]).

According to a previous OECD report, the largest share of over-indebted households is observed among the “middle class” (OECD, 2015[36]).7 Recent data shows that, on average, one in eight households in the middle8 of the income distribution is over-indebted9. Over-indebtedness is particularly prevalent in Chile, the Netherlands and Norway, where it affects at least one in four households. In contrast, the figure is lower than 5% in Austria, Estonia, Germany, Greece, Italy, Latvia, Poland and Slovenia (Figure 4.11).

In most OECD countries, over-indebtedness is more widespread among middle-income households than in the overall population. In Canada, Luxembourg, New Zealand and Portugal, it is at least 5 percentage points higher.

Figure 4.11. One in eight households in the middle of the income distribution is over-indebted

4.5. Conclusions

This chapter has sought to gather evidence on the standard of living of middle-income households in OECD countries and selected emerging economies (Brazil and South Africa) over recent decades. To that end, it has assessed how middle-income households contribute to overall household consumption and allocate spending across consumption categories. It has also examined how the spending of middle-income households has increased relative to income, what consumption categories have driven that increase, and to what extent rising prices and changing preferences have been influencing factors. Finally, it assessed the ability of middle-income households to afford their standard of living, and considered their exposure to financial vulnerability and over-indebtedness.

Middle-income households account for two-thirds of total household consumption expenditure OECD-wide. Except for a few countries, the contribution of middle-income households to overall expenditure is proportionately larger than their shares of income and population. In contrast to OECD countries, in Brazil and South Africa the middle-income class contributes less than upper-income households to overall household consumption.

The ability to save of middle-income households has fallen in the last two decades as expenditure has outstripped incomes, particularly prior to the global financial crisis. It then plummeted between 2007 and 2010 and remained somewhat stagnant between 2010 and 2015.

Housing accounts for an increasing share of middle-income household budgets. Between 1995 and 2015, it increased from a quarter to almost one-third. Expenditure on owned homes made up most of the rise, although spending on rent and utilities also expanded considerably. Rising house prices have been a key factor. In the last two decades, they have grown at twice the rate of inflation and more than one-third faster than median incomes.

Private healthcare spending is further squeezing the budgets of middle-income households, with strong growth in spending on out-of-pocket and private health insurance. In a few countries, the rising costs of higher education also threaten the ability of middle-income families to send their children to university, as prices of education-related goods and services have increased faster than inflation and median incomes.

In addition to rising prices, changes in consumption preferences can explain some of the shifts in middle-income spending. Growing demand for good public services and the gentrification of urban areas have also nudged middle-class households into more expensive segments of the housing market. Similarly, middle-income spending patterns may have been affected by changes in income inequality and how high-income households spend their budgets.

With the rise in the cost of living and expenses growing faster than incomes, many middle-income households find themselves struggling to make ends meet. Some have become financially vulnerable and others spend more than they earn. Moreover, over-indebtedness among middle-income households has become a growing phenomenon in a number of countries.

References

[13] ABS (2017), Household Expenditure Survey, Australia: Summary of Results, 2015-16, Australian Bureau of Statistics, http://www.abs.gov.au/ausstats/[email protected]/Latestproducts/6530.0Main%20Features32015-16?opendocument&tabname=Summary&prodno=6530.0&issue=2015-16&num=&view= (accessed on 1 December 2017).

[19] Alfageme, A. (2019), “El negocio de la sanidad privada se dispara tras los recortes en el sistema público”, El País, https://elpais.com/sociedad/2019/01/28/actualidad/1548704969_178978.html (accessed on 3 March 2019).

[40] Allen, J. et al. (2008), Housing and welfare in Southern Europe, John Wiley & Sons, https://onlinelibrary.wiley.com/doi/book/10.1002/9780470757536.

[7] André, C. (2016), “Household debt in OECD countries: Stylised facts and policy issues”, OECD Economics Department Working Papers, No. 1277, OECD Publishing, Paris, https://dx.doi.org/10.1787/5jm3xgtkk1f2-en.

[38] Balestra, C. and R. Tonkin (2018), “Inequalities in household wealth across OECD countries: Evidence from the OECD Wealth Distribution Database”, OECD Statistics Working Papers, No. 2018/01, OECD Publishing, Paris, https://dx.doi.org/10.1787/7e1bf673-en.

[32] Bertrand, M. and A. Morse (2016), “Trickle-Down Consumption”, Review of Economics and Statistics, Vol. 98/5, pp. 863-879, https://doi.org/10.1162/REST_a_00613.

[21] Blank, R. (2010), Middle class in America, U.S. Department of Commerce, Economics and Statistics Administration, http://www.esa.doc.gov/sites/default/files/middleclassreport.pdf.

[14] Cardoso, M. and M. Montañez (2017), Caída de la desigualdad en consumo: El caso de España durante la crisis, BBVA Research, https://www.bbvaresearch.com/wp-content/uploads/2017/10/Caida-de-la-desigualdad-en-consumo-El-caso-de-Espana-durante-la-crisis.pdf.

[27] Chapple, K. (2016), “Income Inequality and Urban Displacement”, New Labor Forum, Vol. 26/1, pp. 84-93, https://doi.org/10.1177/1095796016682018.

[26] Combes, P. and L. Gobillon (2015), “The Empirics of Agglomeration Economies”, in Handbook of Regional and Urban Economics, Elsevier, https://doi.org/10.1016/b978-0-444-59517-1.00005-2.

[10] Crossley, T., H. Low and C. O’Dea (2013), “Household Consumption through Recent Recessions*”, Fiscal Studies, Vol. 34/2, pp. 203-229, https://doi.org/10.1111/j.1475-5890.2013.12003.x.

[33] Currid-Halkett, E. (2017), The sum of small things : a theory of the aspirational class, https://press.princeton.edu/titles/10933.html.

[9] De Nardi, M., E. French and D. Benson (2012), Consumption and the Great Recession, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2017759.

[29] Dearden, L. et al. (2017), Higher Education funding in England: past, present and options for the future, Institute for Fiscal Studies, https://doi.org/10.1920/bn.ifs.2017.bn0211.

[24] Engsted, T., S. Hviid and T. Pedersen (2016), “Explosive bubbles in house prices? Evidence from the OECD countries”, Journal of International Financial Markets, Institutions and Money, Vol. 40, pp. 14-25, https://doi.org/10.1016/j.intfin.2015.07.006.

[5] Eurostat (2019), Experimental statistics — Overview, https://ec.europa.eu/eurostat/web/experimental-statistics/.

[30] Frank, R. (2013), Falling behind : how rising inequality harms the middle class, University of California Press.

[31] Frank, R., A. Levine and O. Dijk (2014), “Expenditure Cascades”, Review of Behavioral Economics, Vol. 1/1-2, pp. 55-73, https://doi.org/10.1561/105.00000003.

[25] Glaeser, E. (2010), Agglomeration Economics, University of Chicago Press, https://doi.org/10.7208/chicago/9780226297927.001.0001.

[28] Hamilton, M. (2014), Urban and Regional Policies for Metropolitan Livability, Routledge, https://doi.org/10.4324/9781315698434.

[41] Hegedus, J., M. Lux and N. Teller (2013), Social housing in transition countries, Routledge, https://www.routledge.com/Social-Housing-in-Transition-Countries/Hegedus-Lux-Teller/p/book/9780415890144.

[22] Les Echos (2018), “Immobilier ancien : vers une année 2018 record”, Les Echos, https://www.lesechos.fr/10/10/2018/lesechos.fr/0302377357037_immobilier-ancien---vers-une-annee-2018-record.htm (accessed on 12 March 2019).

[6] OECD (2019), Affordable Housing Database, http://www.oecd.org/social/affordable-housing-database.htm (accessed on 28 February 2019).

[15] OECD (2019), OECD Economic Surveys: Hungary 2019, OECD Publishing, Paris, https://dx.doi.org/10.1787/eco_surveys-hun-2019-en.

[16] OECD (2018), OECD Environmental Performance Reviews: Hungary 2018, OECD Environmental Performance Reviews, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264298613-en.

[3] OECD (2018), “The need to follow up on the Stiglitz-Sen-Fitoussi Commission”, in Beyond GDP: Measuring What Counts for Economic and Social Performance, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264307292-6-en.

[20] OECD (2017), Health at a Glance 2017: OECD Indicators, OECD Publishing, Paris, https://dx.doi.org/10.1787/health_glance-2017-en.

[36] OECD (2015), In It Together: Why Less Inequality Benefits All, OECD Publishing, Paris, https://doi.org/10.1787/9789264235120-en.

[1] OECD (2013), OECD Framework for Statistics on the Distribution of Household Income, Consumption and Wealth, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264194830-en.

[17] OECD (2011), OECD Economic Surveys: Czech Republic 2011, OECD Publishing, Paris, https://dx.doi.org/10.1787/eco_surveys-cze-2011-en.

[39] OECD/European Observatory on Health Systems and Policies (2017), Netherlands: Country Health Profile 2017, State of Health in the EU, OECD Publishing, Paris/European Observatory on Health Systems and Policies, Brussels, https://dx.doi.org/10.1787/9789264283503-en.

[34] Petev, I. (2013), “The Association of Social Class and Lifestyles”, American Sociological Review, Vol. 78/4, pp. 633-661, https://doi.org/10.1177/0003122413491963.

[11] Petev, I., L. Pistaferri and I. Eksten (2011), “Consumption and the Great Recession: an analysis of trends, perceptions, and distributional effects”, https://www.tau.ac.il/~itaysap/cons_recess_August_2011.pdf.

[12] PEW (2016), Household Expenditures and Income, Balancing family finances in today’s economy, The Pew Charitable Trusts, http://www.pewtrusts.org/en/research-and-analysis/issue-briefs/2016/03/household-expenditures-and-income.

[18] Schanzenbach, D. et al. (2016), “Where does all the money go: Shifts in household spending over the past 30 years”, Brookings Institution, The Hamilton Project., http://www.hamiltonproject.org/assets/files/where_does_all_the_money_go.pdf.

[2] Stiglitz, J., A. Sen and J. Fitoussi (2009), Report by the Commission on the Measurement of Economic Performance and Social Progress, https://ec.europa.eu/eurostat/documents/118025/118123/Fitoussi+Commission+report.

[37] Tsenkova, S. (ed.) (2009), Housing Policy Reforms in Post Socialist Europe, Physica-Verlag HD, Heidelberg, https://doi.org/10.1007/978-3-7908-2115-4.

[4] United Nations (2018), “Classification of Individual Consumption According to Purpose (COICOP) 2018”, No. ST/ESA/STAT/SER.M/99, Department of Economic and Social Affairs, Statistics Division, https://unstats.un.org/unsd/classifications/business-trade/desc/COICOP_english/COICOP_2018_-_pre-edited_white_cover_version_-_2018-12-26.pdf.

[8] Whitehead, C. and P. Williams (2017), “Changes in the regulation and control of mortgage markets and access to owner-occupation among younger households”, OECD Social, Employment and Migration Working Papers, No. 196, OECD Publishing, Paris, https://dx.doi.org/10.1787/e16ab00e-en.

[23] Wijburg, G. and M. Aalbers (2017), “The alternative financialization of the German housing market”, Housing Studies, Vol. 32/7, pp. 968-989, https://doi.org/10.1080/02673037.2017.1291917.

[35] Wolff, E. (2014), “Household wealth trends in the United States, 1983-2010”, Oxford Review of Economic Policy, Vol. 30/1, pp. 21-43, https://doi.org/10.1093/oxrep/gru001.

Annex 4.A. Data sources

Annex Table 4.A.1. Categories of Classification of Individual Consumption according to Purpose (COICOP)
Code	Description
01	Food and non-alcoholic beverages
02	Alcoholic beverages, tobacco and narcotics
03	Clothing and footwear
04	Housing, water, electricity, gas and other fuels
041	Actual rentals for housing
042	Imputed rentals for housing
043	Maintenance and repair of the dwelling
044	Water supply and miscellaneous services relating to the dwelling
045	Electricity, gas and other fuels
05	Furnishings, household equipment and routine household maintenance
06	Health *
07	Transport
08	Communication
09	Recreation and culture
10	Education
11	Restaurants and hotels
12	Miscellaneous goods and services
1253	Insurance related to health *
Note: * Data used in the analysis for this report includes insurance related to health (COICOP code 1253) as part of health consumption expenditure.
Source: United Nations, Department of Economic and Social Affairs (United Nations, 2018[4]).

Annex Table 4.A.2. Data sources
	Source	Organisation undertaking the survey	Frequency of collection	Years used in the analysis	Imputed years	Sample size (number of households)
Belgium	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 7 200
Brazil	Pesquisa de Orçamentos Familiares	Instituto Brasileiro de Geografia e Estatística	Every 5 years	2002, 2009	-	~ 50 000
Chile	Encuesta de Presupuestos Familiares	Instituto Nacional de Estadisticas	Every 5 years	1996, 2010, 2016	-	~ 15 100
Denmark	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 2 500
Estonia	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 3 600
Finland	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 3 600
France	Household Budget Surveys^a	Institut national de la statistique et des études économiques	Every 5 years	2006, 2011	-	~ 15 600
Greece	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 3 500
Hungary	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 9 900
Ireland	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 5 800
Lithuania	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 6 100
Latvia	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 3 700
Mexico	Encuesta Nacional de Ingresos y Gastos de los Hogares	Instituto Nacional de Estadística y Geografía	Every 2 years	2000-2016	-	~ 70 300
Poland	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 37 000
Portugal	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 9 500
Slovak Republic	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 6 100
Slovenia	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 3 900
Spain	Encuesta de Presupuestos Familiares^a	Instituto Nacional de Estadisticas	Annually	2006-2015	-	~ 22 000
	Household Budget Surveys^a	Eurostat	Every 5 years		1998, 2005^b	~ 22 000
South Africa	Income and Expenditure Survey	Statistics South Africa	Every 5 years	2000-2010	-	~ 24 600
Sweden	Household Budget Surveys^a	Eurostat	Every 5 years	2010	1998, 2005, 2015^b	~ 2 000
United Kingdom	Food and Living Conditions Survey	Office for National Statistics	Annually	2002-2014	-	~ 5 100
United States	Consumer Expenditure Survey	Bureau of Labour Statistics	Annually	2005-2016	-	~31 000
Note: a) Incomes have been adjusted for non-reporting using experimental statistics on income published by Eurostat (https://ec.europa.eu/eurostat/web/experimental-statistics/income-consumption-and-wealth). b) Data for these years have been imputed using tabulations published by Eurostat and micro-data from 2010. Expenditure has been imputed taking into account the growth rates in expenditure by quintile of disposable income.

Notes

← 1. In line with the standard economic approach and other chapters in the report, this chapter considers the “middle-income class” to be people living in households with equivalised disposable incomes ranging between 75% and 200% of the national median. Households with incomes of below 75% are “lower income” and those whose incomes exceed 200% are “upper income” (see Box 2.1 in Chapter 2).

← 2. On average across OECD countries, homeowners make up about 70% of middle-income households – outright owners are about 40% and owners with mortgage almost 30% (OECD, 2019[6]). Outright home ownership is widespread in Mexico and the countries of Central, Eastern and Southern Europe. In Central and Eastern Europe, after the fall of the communist regimes, tenants could buy the dwellings in which they lived at a low price (Hegedus, Lux and Teller, 2013[41]; Tsenkova, 2009[37]). In Mexico and Southern European countries, mortgage markets started developing only recently and families have traditionally a strong role in facilitating home ownership through inheritance or financial support (Allen et al., 2008[40]).

← 3. In most Central and Eastern European countries, the share of tenant households is generally very low. In Latvia, 13% of households are tenants and one-third of them have subsidised rents.

← 4. People who live or work in the Netherlands are legally obliged to take out standard private health insurance. Insurers must accept all applicants and are expected to contract providers on the strength of quality and price (OECD/European Observatory on Health Systems and Policies, 2017[39]).

← 5. Over the long term, housing prices can be fairly well tracked by fundamentals. In the short term, however, housing prices may overshoot due to extrapolative expectations, in some case creating huge bubbles (André, 2016[7]).

← 6. ^.Inconspicuous consumption reveals class position while not intending to be socially or culturally visible. It includes goods and services such as gardeners, childcare, car services or education and retirement plans (Currid-Halkett, 2017[33]).

← 7. The term “middle class” was used in the report to refer to the three middle quintiles of the income distribution (i.e. the middle 60% the population).

← 8. The term “middle of the income distribution” refers to the middle quintile (i.e. the 20% of the population in the middle of the income distribution). This definition differs from the “middle-income” used throughout the chapter. It was necessary because the OECD Wealth Distribution Database collects data by income quintiles. While covering a narrower group of people, the middle quintile only includes households that are also part of the “middle-income”, in contrast to other definitions such as “the middle 60%”.

← 9. Over-indebtedness is defined as having a debt-to-income ratio of over 3 (Balestra and Tonkin, 2018[38]).

End of the section – Back to iLibrary publication page