Chapter 4. Households and their financial behaviour1

Michael Andreasch

(Oesterreichische Nationalbank)

This chapter explains the financial behaviour of households as consumers, investors in financial and non-financial assets (dwellings), and owners of unincorporated enterprises. At the aggregate level, households are typically net “savers” and net lenders. This means that they provide funds to other domestic sectors and to non-residents. The creation and the structure of households’ financial wealth depends on several factors, such as the level of and change in disposable income and saving, the financial system, market conditions, and regulations related to pensions and taxes. To fully comprehend the relevant data, it is also important to understand the measurement of the relevant statistics, and to clarify what is actually meant with “households”. Data sources and compilation routines as well as the statistical standards of the 2008 System of National Accounts (2008 SNA) are addressed in this chapter, with the aim to show which conclusions can be drawn from the data – and which cannot.

1. Households as consumers, entrepreneurs, investors and net lenders

Components of GDP, such as final consumption, investment and net exports/imports are seen as the main drivers of economic growth. Households are involved in this process, directly and indirectly, through their final consumption expenditures and through government consumption on behalf of households. Households also act as entrepreneurs by investing in buildings, machinery and equipment related to their business as self-employed workers or sole proprietors. As owners of dwellings they also invest in real estate. All these activities are reflected in the non-financial accounts, which have a direct link with the financial accounts and balance sheets. The sequence of non-financial accounts provides information on saving (i.e., disposable income minus final consumption), which is a key element for the provision of funds. This saving is partly used for investments in non-financial assets. For the remaining part, households are net lenders, which is the case in most OECD countries. In some countries, however, the investments in non-financial assets are larger than saving, and households on an aggregate level need additional funding from other sectors, thus becoming net borrowers. The resulting net lending/borrowing is mirrored in the financial accounts. This parity can be written as follows:

Saving plus Net capital transfers minus Non-financial investments = Net lending/borrowing = Financial investments minus Financing (in the case of households, these are primarily debt liabilities)

After a short discussion of compiling with estimates for the transactions and positions of households, the financial behaviour of households is discussed in much more detail below. However, before doing so, one should note that the term “households” often includes “non-profit institutions serving households (NPISHs)”. These institutions consist of, for example, churches or religious societies, social, cultural, recreational or sports clubs, political parties, trade unions, etc. Although culturally important, they are less significant from a purely economic perspective. Usually, little financial information is available on these institutions, which is why they are often grouped together with households.

Measuring transactions and positions of households

From a statistical point of view, it is important to stress that the data on households are often derived from a horizontal balancing procedure, in which the statistics on households are based on the information of counterparty sectors. For example, in the case of debt, the liabilities of households are estimated through the use of information from the creditors. These creditors traditionally consist of domestic banks and other financial institutions, and also the government (granting subsidised loans). This procedure is quite common due to the general lack of timely data with the requested frequency from households themselves. Moreover, results from household surveys often show a lack of coverage, due to, for example, an underrepresentation of the rich and wealthy (as they are less inclined to participate in a survey). As compared to a survey of the average age of people, where missing the extremes will have a negligible impact on the results, missing extremely wealthy people can have a very significant impact on the aggregate results. Therefore, from a statistical perspective, fully capturing the financial and non-financial behaviour of households via surveys is not straightforward, whereas the counterparty information is usually well defined and observed; see further discussion of data sources in the “Going further” section at the end of this chapter.

Role of households

Household behaviour, both financial and non-financial, is a key element of national accounts, due to the role households play in the generation of income and wealth of the economy. One could also say that, ultimately, it’s all about households, in the sense that the economy is a means for people (or households) to generate income and well-being.

The financial accounts and balance sheets cover the accumulation of household financial assets and debt as well as the resulting stocks. The household sector simultaneously combines a variety of economic activities: consumers of goods and services; small business entrepreneurs investing in non-financial assets, farmers or other persons running sole proprietorships; stakeholders in companies; owners of dwellings; and finally net lenders or net borrowers. Individuals in the household sector are extremely heterogeneous in terms of the level and structure of (financial and non-financial) wealth and therefore in their (financial) behaviour, which means that monitoring and analysing at the aggregate level only tells part of the story.

Households as consumers

Households have two distinct sources of financing their final consumption expenditures: internal financial sources, which mainly consist of their disposable income (including net capital transfers), and external financial sources, such as consumer loans to supplement their internal sources. How, and how much, households consume critically affects economic growth. The contribution of private consumption to Gross Domestic Product (GDP) ranged between 30% and 70% for OECD countries in the years 2011-15. In the majority of countries, the ratio is close to the mean value of 55% for all OECD countries; see Figure 4.1.

In other words, on average, 55 cents of every dollar or euro of total GDP is related to final consumption expenditure of households, such as expenditures on housing (rentals, including imputed rentals for owner-occupied dwellings, water, electricity, gas and other fuels), health, education, furnishing, equipment, food and beverages (including restaurants), etc. In many countries expenditures on health and education are actually done or reimbursed by the government. They are thus part of government consumption expenditure, but in the analysis of the so-called “actual consumption” they are considered as final consumption by households. Most of the funds spent by households are related to consumption of non-durable goods and services. However, consumption also includes expenditures on durable goods, like cars, TVs and computers, unless they are used in the production of goods and services by unincorporated enterprises and self-employed, or they relate to major renovations by owner-occupiers of dwellings.

The most important factor for the analysis of the financial behaviour of households as consumers is the volume of new consumer loans, which are normally granted by banks or other financial institutions. The measurement of new consumer loans is rather challenging, because financial accounts and balance sheets are usually shown on a net basis (deducting down-payments of existing loans from the incurrence of new loans). However, data derived from the ECB interest rate statistics on new loans show that, in the period 2005-15, consumer loans typically account for around 4% of overall consumption in the Euro Area as a whole, with a rather stable pattern over the years. One may assume that most of these consumer loans are related to the purchase of consumer durables such as cars.

Households as investors in financial and non-financial assets (dwellings)

The part of disposable income that is not consumed represents saving. Saving is important for understanding both shorter- and longer-term financial behaviour of households. Saving by households is primarily invested in both financial assets (deposits, securities, pension plans, etc.) and dwellings. Saving may also be used – primarily by self-employed persons – for investments in commercial property, machinery, equipment, etc. needed to set up or extend small businesses. In most countries, the single most important form of household debt relates to mortgage loans for the purchase of dwellings. The development of the debt levels is heavily influenced by the price movements in the housing market.

Figure 4.2 shows household financial investments and non-financial investments of selected OECD countries in the last 20 years. In line with the vulnerabilities after the 2007-09 economic and financial crisis and the low interest rates in the middle of the 2010s, there is a clear shift in preferences from investing in financial assets towards investing in non-financial assets, notably investments in real estate and gold. In 2015, the latest year available, financial investments regain strength, and are almost equal to non-financial investments.

The “life-cycle” and “permanent income” hypothesis, as developed by Franco Modigliani and Milton Friedman, suggests that individuals borrow when they are young (e.g. mortgage loans for the purchase of dwellings, student loans or small business loans), save during the majority of their working lives, and dis-save when they retire, possibly also leaving, in the permanent income hypothesis, bequests to their descendants. In normal circumstances, households as a whole tend to consume less than they earn, thus providing a steady flow of saving to the economy. Economies with few retirees and many young and middle-aged workers will typically save more than economies characterised by a high number of retirees. How demographic trends affect macro-economic developments is discussed in more detail in Chapter 9.

Households as entrepreneurs and owners of companies

Before discussing the financial behaviour of households in their role as entrepreneurs, it is key to understand their role in the production of goods and services, as owners of small businesses or as self-employed. On the one hand households can act as entrepreneurs, and invest in non-financial assets related to their business. These investments are reflected in the non-financial accounts, and therefore do not directly affect the understanding of the financial accounts and balance sheets. On the other hand, households can also operate as stakeholders of companies, by making a portfolio investment, through buying shares on organised stock markets, or investing more strategically, such as by acquiring stakes in small and medium-sized enterprises (SMEs). For some countries, equity holdings in SMEs play an important role in the inter-linkage between households and the non-financial corporations’ sector, depending on where the enterprises are recorded. For more details on the non-financial corporations sector, see Chapter 5. While the role of households as consumers is rather well defined and traceable in the national accounts data, one usually cannot separately distinguish households as entrepreneurs for the full set of accounts, because their entrepreneurial behaviour is fully intertwined with their behaviour as consumers.

Looking more closely to the sector allocation of SMEs, the 2008 System of National Accounts (2008 SNA) makes a distinction between incorporated and unincorporated enterprises. Enterprises with a separate legal status, which includes relatively small limited liability companies, are always recorded in the corporations sector. On the other hand, an unincorporated enterprise is normally recorded in the household sector, unless the enterprise “… has sufficient information to compile a complete set of accounts and is operated as if it were a separate corporation and whose de facto relationship to its owner is that of a corporation to its shareholders” (SNA 2008, paragraph 4.42). The latter guidance may, however, be implemented quite differently across countries. For example, in the National Income and Production Accounts (NIPA) of the United States, all enterprises, both incorporated and unincorporated, are grouped together in the “business sector”. However, for reasons of international comparability, the United States provides data to the OECD in which the enterprises are allocated to the various sectors in line with the international standards.

There are other diverging practices across countries in the allocation of unincorporated enterprises to either the households’ or the corporations’ sector, but not so extreme as in the case of the US NIPA. In some cases, the diverging practices may be related to the kind of source information that is available for the household sector. As stated before, the compilation of data for the households’ sector is often not based on direct observations, but on data from the counterparty sectors, such as the information on lending which may be based on information from banks. In these sources, the delineation of sectors may be different from the one applied in the 2008 SNA. Another point concerns the allocation of family trusts (foundations or funds), which are designed purely to manage the financial and non-financial wealth of very wealthy individuals. If these funds have some form of autonomy, they should be treated as part of the financial corporations’ sector in the SNA, but this rule of autonomy may be applied quite differently. Divergences across countries may also be the result of institutional arrangements, such as differences in legislation affecting the economic behaviour of people across countries. They may also be related to differences in the interpretation and the application of international standards by statisticians across the world, especially when it comes to interpreting when the de facto relationship of an unincorporated enterprise is that of a corporation to its shareholders, as discussed above. Some countries apply this criterion very strictly, while others use a looser interpretation. The analysis of the accounts for households in their role as entrepreneur, including the indicators that can be derived from these accounts, therefore needs to be done with some caution, as these divergences may hamper the international comparability of the results.

For example, Figure 4.3 shows the equity holdings of households for a number of OECD countries, as a percentage of their total wealth. The differences are quite startling, ranging from more than 50% in Estonia to next to zero in the Slovak Republic. This may be related to differences in the prevalence of households to invest in shares and other equity. It may also be related to accumulated pension wealth in some countries that have established large funded pension schemes. However, an important factor can also be related to the recording of a significant number of SMEs as separate entities in the corporations’ sector. If this is the case, it automatically leads to an increase of equity wealth, while in the case of consolidation of unincorporated enterprises in the households’ sector, no equity is registered. In the latter case, instead of equity, the assets and liabilities of the enterprise are recorded in the accounts of the household sector.

To arrive at a better understanding of the role of households as entrepreneurs, it may be necessary to go back to the survey data, and try to collect information on the transactions and positions that are specifically related to the relevant enterprises. This information could, for example, be derived from integrated household surveys which include income statements and balance sheet data of (unincorporated) enterprises. Tax data may be another source of information, as tax authorities often ask specific questions on the income and wealth related to (unincorporated) SMEs. To better understand the role of households as stakeholders of companies, one may again want to look at more granular data. For example, Magri (2009) has tested for Italy the importance of a household’s initial wealth for becoming an entrepreneur. The results showed that the probability of becoming an entrepreneur is positively correlated with a household’s initial wealth, but that the effect is also dependent on the size of wealth, with a smaller effect for the richest households.

Households as net lenders or borrowers

Saving plus net receipts of capital transfers minus the investments in non-financial assets represent net lending (or net borrowing), i.e. the residual income that is available for increasing financial wealth. By definition, this residual is equal to the balance of net purchases of financial assets minus the net incurrence of liabilities, although in practice one may observe statistical discrepancies. From a macro-economic viewpoint, households are typically net lenders, i.e. they have a positive net lending/net borrowing, and thus add each year to their net financial wealth. The total stock of wealth, including non-financial assets, is reflected on the balance sheets. Apart from net acquisitions of non-financial assets and the above-mentioned transactions in financial assets and liabilities, the stocks may also be affected by holding gains (or losses) and by other changes in the volume of assets (e.g. catastrophic losses, theft, etc.); see Box 4.1 for a more detailed explanation.

Box 4.1. Moving from (accumulated) flows to stocks

As noted in Chapter 1, changes in financial stocks are determined by net purchases of financial assets and by the net incurrence of liabilities or financing. These transactions are recorded in the financial accounts. In addition, stock levels may change as a consequence of flows recorded in the “other changes in assets accounts”, which comprise price changes (revaluations) and other changes in volume of assets. The latter are often referred to as stock-flow adjustments. A recap, with a particular focus on aspects relevant for households, is provided in the Table 4.1.

Table 4.1. Summary of changes in stocks
Stocks_t-1	Changes of stocks due to			= Stocks _t
Stocks_t-1	+/-Transactions _t	+/-Revaluations _t	+/-Other changes in volume _t	= Stocks _t
	Gross purchases minus gross sales (including realised holdings gains and losses) plus accrued interest (reinvestment of interest until payment)	Changes in the value of stocks due to changes in the exchange rates of foreign currencies; market price changes based on unrealised holding gains and losses including writedowns of tradable securities; changes due to model assumptions (discount rate, etc.), and price changes in unit-linked contracts and defined contribution schemes of individual life insurance and pensions ; all changes due to price and volatility changes of underlying assets in the case of financial derivatives	Changes due to changes in demographic assumptions in the case of insurance and pension schemes plus Uncompensated seizures, write-offs of debt, methodological changes

On average, stock-flow adjustments account for one-third of the total change in financial wealth of households for all OECD countries in the last 20 years. Net purchases thus account for two-thirds of the increase in household financial wealth. However, as Figure 4.4 clearly shows, patterns are quite different across countries. In many countries, the other changes in assets represent more than half of the changes in the value of stocks.

The resources and uses of funds can be analysed by comparing the total financial and non-financial investments with the underlying funding through saving, net capital transfers and borrowing (for households this mainly represents bank loans); this can provide very useful insights. The analysis adds to the understanding of the possible causality between certain types of investments and the sources of funding, i.e. whether or not saving directly influences financial investments or how non-financial investments are financed through internal and/or external funding. Figures 4.5 to 4.7 below illustrate this analysis for three types of countries; i) countries where households show a rather stable net lending situation; ii) countries where households switched from a role as net lenders to a role as net borrowers role and vice versa; and finally iii) countries where households are more or less permanent net borrowers.

In 14 OECD countries,2 households almost continuously provided funds as net lenders to other sectors in the past 20 years. Generally, net lending as a percentage of gross disposable income (GDI)3 ranged from close to zero to 10%, with an average of 6%, with the exception of Switzerland, where households showed consistent net lending of more than 10%. However, quite remarkable differences can be observed over time. While households’ net lending in Sweden was close to zero in the late 1990s, their net lending increased in recent years to more than 10%. In contrast, the net lending position of Italian households decreased steadily from above 10% in 1995 to 3% in 2013-14. The stocks of accumulated net financial wealth, not taking into account non-financial assets, for these countries ranged from 1.1 to 4.0 times their GDI in 2014. Figure 4.5 illustrates trends in net lending for a selection of these 14 countries.

In another nine countries,4 one can observe a shift from households being net lenders to becoming net borrowers, or vice versa. In most of these countries, households were net lenders until the beginning of the 21th century (with the exception of Irish households), and they became net borrowers against the background of the burst of the internet bubble. In most of the countries, the financial position changed again into a net lending position after the 2007-09 economic and financial crisis, partly due to an increase in the saving ratio. For example, in Estonia and Ireland the gross saving ratio was negative until 2007 and became positive from 2008 onwards. Net financial wealth in relation to (adjusted) GDI deviates significantly across these countries, ranging from 0.6 (Norway, Slovak Republic) to 3.1 (the United Kingdom) and 4.0 (the United States). For a further discussion on the impact of financial crises and globalisation on key financial accounts indicators, reference is made to Chapter 10. Figure 4.6 illustrates trends in net lending for a selection of these nine countries.

Finally, in five countries, including three Northern European countries, households were mainly net borrowers in the last 20 years, with a downward trend since the beginning of the century up until the 2007-09 economic and financial crisis. All these countries had negative saving ratios or saving ratios close to zero, well below the average of OECD countries for the entire period. Contrary to the households which showed a change in behaviour, they did not increase their saving after the financial crises. The best example to show that nearly permanent net borrowing does not automatically lead to a lower net wealth position is Denmark. Net financial wealth, again excluding non-financial assets, stood at 3.1 times their (adjusted) GDI at the end of 2014. This ratio is above the average value for all OECD countries.

Looking at these developments from the perspective of the counterparty sectors, changes in the financial position of households can also have an impact on the financial position of other economic sectors, as household net lending can be an important source of funding for other domestic sectors or the Rest of the World. On the other hand, if households are in a net borrowing situation, they may become very dependent on funding inflows through the financial system (mainly banks).

2. The interlinkages between income, saving, financial and non-financial investments, and debt

It is evident that the accumulated stock of financial assets can be important for individual well-being, as that stock represents resources available to maintain adequate levels of consumption and welfare, including, for example, after retirement. It also makes households more resilient to deal with misfortune, and thus provides them with more security. Households’ financial assets are the result of two factors: the net acquisition of (certain types of) financial instruments from income not spent, and the impact of changes in the prices of the assets accumulated in the past. Overall households are free in their decision whether to invest in financial or non-financial assets. However, two aspects have to be taken into consideration here. First, individual households in countries with a high degree of private (funded) pension schemes are usually obliged to invest in pension plans, typically by building up pension entitlements held with pension funds and insurance corporations (see also Chapter 9). Second, households with outstanding bullet loans will have to invest in repayment vehicles, usually saving deposits, (unit linked) life insurance contracts or equity-oriented investment fund shares.

Looking at national accounts aggregates, the main questions that may arise during the analysis of households’ financial investments, borrowing and the accumulation of financial (net) wealth are: “What are the main drivers of the developments in the different components?” and “How are they interlinked?” The following issues may be important in determining the actual financial behaviour of households over time, and allowing us to find explanations for the differences in the level and structure of (financial) wealth and liabilities across countries:

From a macro-economic viewpoint:
- the level of and the development in disposable income and saving;
- the arrangements of the financial system (and the direct and indirect participation of households);
- changes in interest rates, and changes in prices on the stock markets and the housing market;
- the presence of funded pension schemes and social security systems;
- investments in and stocks of non-financial wealth (especially housing wealth);
- the tax regime for (the income on) assets and (payments of interest on) debt; and
- various legal aspects, such as foreign exchange liberalisation and credit standards.
From a microeconomic viewpoint:
- the level of and development in income and wealth of individuals; and
- the impact of the life cycle hypothesis.

Figure 4.8 provides a graphical representation of the financial accounts, and the interconnectedness between the different stages of the sequence of accounts. A distinction is made between i) the direct impact (marked with solid arrows) of certain elements of income, saving, financial and non-financial investments, borrowing, stocks of financial and non-financial assets, and debt; and ii) the analytical interlinkages (marked with dashed arrows) between these components in the sequence of accounts.

In Figure 4.8, the presentation of accounts starts with disposable income. Disposable income is the income available to households to purchase consumer goods and services. It consists of the income received (e.g. compensation of employees, net income from engaging in unincorporated enterprises, interest, dividends, social benefits received, etc.) minus payments (e.g. taxes and social contributions, interest on mortgage loans, etc.). The result of disposable income minus final consumption expenditures, including the addition of the so-called “adjustments for the change in equity of households in pension funds”, which corresponds to the financial investments in pension entitlements, represents the saving of households. The saving rate, the total saving as a percentage of disposable income,5 is a key element in the analysis of the financial behaviour of households.

Saving represents the ability of households to make financial and non-financial investments. Here, it should be made clear that saving, as the term is used in national accounts, is not equivalent to putting additional money on saving deposits, which is only one way of investing money which is available through saving. Saving, including net receipts of capital transfers (like inheritances and large one-off donations between households) and the (net) acquisition of non-financial assets, is part of the so-called “capital account” in the sequence of accounts.

Financial investments, typically in the form of changes in cash holdings and current accounts, investments in saving deposits, net purchases of bonds, shares and mutual fund shares, and investments in other forms of participation like the holding of equity stakes in limited liability companies, and additional borrowing, usually in the form of loans, are recorded in the financial account. The capital account and the financial account both form part of the so-called “accumulation accounts”, i.e. all the accounts which record transactions that have a direct impact on one or more items on the balance sheets. The other set of accounts in the national accounts are the so-called “current accounts”, which consist of the production account, the generation and (re)distribution of income accounts, and the use of income account. For more details, see Chapter 1 and the OECD publication “Understanding National Accounts” (2014).

Financial investments on the asset side and changes in borrowing on the liability side are major drivers for the increase or decrease of stocks of financial assets and liabilities. Investments in housing in the form of dwellings (including land) significantly contribute to the change in non-financial assets owned by households. In addition, revaluation effects, and other changes in volume like catastrophic losses and reclassifications, also add to the changes in balance sheet positions. In this respect, it should be noted that national accounts do not consider holding gains (or losses) as part of property income, in contrast to the concept of “yield”, which might influence the decision-making process of households in their portfolio allocation. Nevertheless, the levels of both financial assets and debt have an indirect impact on the level of disposable income, due to the fact that property incomes derived from financial assets minus interest paid for debt add to disposable income. All the above details and interlinkages are provided by a full-fledged system of national accounts. More details on the whole sequence of accounts is presented in Annex 2 of the 2008 SNA.

An alternative to the above, more traditional presentation in the form of a sequence of accounts in Figure 4.8, is represented in Figure 4.9. It shows the total “resources and uses of funds”. The key difference is that the national account term “saving” is not used. What is presented is an overview of all available (financial) resources: i) internal financing in the form of disposable income, including the adjustment for the change in equity of households in pension entitlements and net capital transfers, and ii) external financing (normally in the form of loans and trade credits, broken down by purpose to reflect the linkage to certain items in the use of funds). These resources are then mirrored by the total uses, broken down into final consumption expenditures and non-financial and financial investments.

Consequently, this type of presentation sums up final consumption (above the line “saving” in Figure 4.8) and non-financial investments (below the line “saving” in Figure 4.8). Both elements are relevant for national accounts in different stages of the sequence of accounts, but the distinction is less relevant for individuals in their economic behaviour. The purchase of a new car (classified as final consumption) is often treated by an individual in much the same manner as the renovation of a dwelling (classified as non-financial investments), using both disposable income and debt financing as the resources for these expenditures.

Regardless of which concept is used, one should realise that there are always limitations to the analysis of the financial behaviour of the household sector. Figure 4.10 illustrates the interconnectedness between disposable income, saving and net financial wealth in a cross-country comparison (Andreasch et al., 2010), by showing data for the savings ratio, the disposable income per capita, and the net financial wealth per capita, as the arithmetic mean for the years 2005-09. The results highlight that there is no such thing as a straightforward link between the saving ratio and net financial wealth. Although the majority of households in this exercise had a disposable income between 20 000 and 26 000 EUR per capita, the savings ratios and the net financial assets are very different across countries.

Furthermore, it is important to look at the linkages between saving and (financial and non-financial) investments, as well as the relationship between internal financing in the form of saving and external financing in the form of loans. There can be a close interaction between (the structure of) financial wealth and the volume of non-financial assets, especially when it comes to housing wealth. The “Three-Equation-Model” (Turk, 2015), for example, evaluates the interaction between house prices, housing stocks and household debt. Moreover, the increase (decrease) of wealth due to the accumulation of assets and holding gains can have a significant impact on household final consumption expenditure, which is a key element in economic growth. Recent studies show a large amount of research and policy analysis on these issues, thus putting even more emphasis on the importance of having high quality data on financial accounts and balance sheets.

The top half of Figure 4.11 illustrates the levels of financial assets of households as a percentage of their (gross) disposable income in OECD countries, while the bottom half shows the composition of the stocks of financial assets. The level of total financial assets diverged strongly across OECD countries. In the Netherlands, the ratio was 6.6 times disposable income in 2014, compared with 1.2 times disposable income in the Slovak Republic. Looking at the composition of financial assets, it can be observed that countries with a high proportion of funded pension schemes and life insurance contracts (including the so-called third pillar pension plans) are usually ranked above the median. In contrast, countries with households which tend to invest a major part of their financial portfolio in currency and deposits tend to rank below the median. In this respect Japan is one of the exceptions. Households in Japan tend to invest every second yen in deposits, but notwithstanding this behaviour their financial wealth to disposable income ratio was 4.4 in 2014, above the median financial assets to disposable income ratio for OECD countries in 2014 (3.6).

Many have researched the impact of the degree of convergence of financial systems across countries, and the relationship between financial systems and the structure of household financial wealth, against the background of a growing liberalisation and globalisation in the 1990s and 2000s. One would expect that globalisation, deregulation, economic integration, and harmonisation of regulations and corporate governance rules would support the harmonisation of financial system characteristics across countries. At a minimum, the increased interest of researchers into these types of issues has raised the question of whether the traditional distinction between the “bank-based” European continental financial system, with its focus on deposits, versus the “market-based” Anglo-Saxon financial system, with its focus on shares and other types of equity, is here to stay. In any case, convergence is something that needs to be studied over a longer period of time. De Bonis et al. (2012) analysed the developments across nine OECD countries (Austria, Canada, France, Germany, Italy, Japan, Spain, the United Kingdom and the United States) for the period from 1980 to 2005, and showed that there is certain evidence of international convergence for households’ total financial assets, equity components, insurance reserves and pension entitlements, while there are rather mixed results, and often no convergence, for currency, deposits and debt securities. The intensity of bank intermediation, in particular, was quite different across the analysed OECD countries.

De Bonis’ analysis of long-term developments can also be decomposed by time periods. In the 1980s, which were characterised by the smooth growth of financial assets against the background of a credit ceiling, high bank reserve requirements and certain limitations to capital movements, constrained the expansion of financial assets in nearly all major OECD countries. This period was followed by the so-called “dot.com” bubble and the subsequent burst at the beginning of the 21st century. The boom resulted in a strong increase in stock market prices of tradable shares and mutual fund shares in households’ portfolios. Subsequently, the 2007-09 economic and financial crisis caused financial (net) wealth to contract in the majority of countries. More pronounced effects could be observed in countries with high levels of investments in equity and/or countries in which pension entitlements related to defined contribution pension schemes (where the value of the entitlements has a direct link with the performance of the accumulated assets) were prominent, like the United Kingdom or the United States. The 2007-09 economic and financial crisis exposed the vulnerabilities of the financial situation of banks and other financial institutions, and led to an increase in government debt and a more pronounced focus on private debt. More details on the impact of the crisis are provided in Chapter 10.

Household financial behaviour in an ageing society

Against the background of an ageing population, a key factor driving long-term convergence in financial behaviour may be related to the organisation of employment-related pensions. Continental European countries, which currently primarily run pay-as-you-go social security systems, tend to increase, with the support of government incentives, the establishment and/or growth of the “second pillar” of funded collective pension schemes and the “third pillar” of individual pension arrangements (via life insurance or other forms of saving). Data for OECD countries show that funded pension entitlements in twelve countries account for 90% of total pension fund assets of all OECD countries. Within this group, the United States represents more than half of the pension fund assets of all OECD countries (Van der Wal, 2014).

The changing demographic conditions clearly raise challenges in ensuring the long-term adequacy of retirement benefits as well as questions about the sustainability of future government expenditures on pensions. The transformation from unfunded social security schemes into funded pensions may be accompanied by risk transfers to households, which could increase the vulnerability of the financial wealth of households, especially if the pension entitlements are directly related to the investment returns on accumulated assets. It has been observed, for example, that households’ pension entitlements in countries with a focus on funded pension schemes dropped relatively strongly during the 2007-09 economic and financial crisis. This volatility may have a detrimental impact on future final consumption and saving streams, thus affecting the whole economy. Here, it should be noted that the recording and measurement of pension entitlements is not without its own problems, certainly if one would like to compare household wealth across countries. Further explanations are provided in Chapter 9.

More generally, when it comes to preparing for retirement, individuals can earmark nearly all financial instruments (cash, deposits, tradeable securities, and individual and collective pension plans) as well as their housing wealth, in addition to income streams from pay-as-you-go social security systems, to sustain their consumption level after retirement. The pension entitlements that are currently recorded in the system of national accounts only relate to (partially) funded employment-related pension schemes. Furthermore, entitlements from individual life insurance are recorded as “life insurance and annuity entitlements”. On the other hand, entitlements that can be derived from pay-as-you-go schemes are not recognised as assets. This clearly hampers international comparability, as households may consider entitlements derived from the latter schemes as strong as the entitlements from funded pension schemes.

According to the requirements of the 2008 SNA, funded pension entitlements are split into defined benefit (DB) schemes and defined contribution (DC) schemes. In DC schemes, the pension entitlements are equal to the assets accumulated through past payments of contributions and investment income earned on the assets (after deduction of the costs of pension funds). The entitlements thus follow closely the market values of the accumulated assets. As pension funds normally apply fair value pricing with little variation in valuation methods across countries, the relevant pension entitlements are comparable across countries offering such a system, like Australia and the United States. In contrast, in the case of DB-systems, such as those in Canada, the Netherlands, Switzerland and the United Kingdom, pension entitlements are based on average wages and the number of years that someone has contributed to the scheme. The entitlements are calculated as the net present value of future benefits, appropriately discounted. However, the methods used to calculate entitlements in the various countries can differ considerably, thus affecting international comparability. For example, the discount rate applied has a significant impact on the level of these entitlements. In some countries this may be a market interest rate, whereas in other countries a long-term interest rate of 3-4% is applied. DB pension schemes may be under- or overfunded, and there is no direct link between the value of pension entitlements and the accumulated assets. It is assumed that the sponsoring employer will take care of the shortfall in funding.

To allow for a better international comparability of pension entitlements, a supplementary table has been developed as part of the 2008 SNA, with the aim of showing a comprehensive picture of all pension schemes, including the contingent entitlements related to pay-as-you-go social security schemes. As discussed earlier, a key question in measuring the level of pension entitlements for DB pension schemes (and, more broadly, pay-as-you-go schemes) is related to the discount rate used to arrive at the current value of future retirement benefits. If countries used the so-called Freiburg model, with the key assumption of a real discount rate of 3% (baseline), the entitlements for unfunded government employees’ schemes would increase the assets of households by roughly 50% of GDP, while the entitlements for social security pay-as-you-go schemes would boost the level of assets on average by more than 230% of GDP. At the same time, these pension entitlements would have an equivalent impact on the level of government (contingent) liabilities.

The main counterparties of households

The financial behaviour of households can be monitored by looking at the changes in the types of financial assets they invest in, and the types of debt they incur. But it may also be useful to monitor changes in the debtors of household assets and the creditors of household debt. To be able to apply such an analysis, there is a need to go beyond the traditional financial accounts and balance sheets, and analyse the interlinkages between the various sectors. The 2007-09 economic and financial crisis showed the importance of risks related to the interconnectedness between countries and sectors. As a reaction to this, more countries started compiling from-whom-to-whom (FWTW) matrices, presenting, for each financial instrument, which sector invests in and borrows from which other sector. As discussed in previous chapters, this allows for an assessment of the importance of specific interconnections, e.g. the importance of households investing in bank deposits and bank debt securities for the refinancing position of banks. For illustrative purposes, the results for Austria are shown in Figure 4.12. The thickness and the direction of the arrows indicate the importance of certain creditors for the liability side of the debtors, whereby all financial instruments have been taken together. The size of the circles indicates the importance of financing within the same sector. One of the results that can be derived from Figure 4.12 is that households and non-resident investors are the main financers of the domestic monetary financial institutions (i.e. the traditional banks), with both contributing nearly the same proportion.

A more detailed analysis of household debt

Data derived from the financial accounts and balance sheets can also be used for a more detailed analysis of household debt. This debt mainly consists of loans (often granted by banks and other financial intermediaries) for housing (purchase and renovation), final consumption, and investments in non-residential buildings and equipment (mainly by entrepreneurs as part of the household sector). From the available time series data, it is evident that in OECD countries, on average two-thirds of household debt (ranging from 53% to 83% for individual countries) consist of housing loans. The analyses of household financial behaviour usually look at the household sector as a whole, thus ignoring the fact that not all households are indebted, and also that not all households have housing loans despite owning a dwelling or other kinds of real estate wealth. To arrive at a better understanding of potential risks and vulnerabilities, it is important to have more details on the various subgroups of households, the compilation of which is in a stage of development at the time of writing this book;6 see also the next section of this chapter.

Figure 4.13 illustrates the indebtedness of households in relation to their GDI in 2014, a common presentation which can also be found in the OECD Dashboard on Households’ Economic Well-being.7 The debt service ratio (defined as the ratio of interest payments plus amortisations to GDI8) shows the proportion of the income which cannot be used for consumption. Denmark and the Netherlands show the highest debt levels, which is also driven by tax incentives on mortgage loans. Consequently, the debt service ratio for these countries is the highest amongst the observed OECD countries. Households in Germany, Austria and Italy had debt levels below 90 per cent of GDI, and less than 7 per cent of their GDI “reserved” for debt service.

The Three Equation Model

The so-called “Three Equation Model”, as shown in Figure 4.14, is often used to explain the interlinkage between i) residential investment in housing, ii) house prices, and iii) housing debt, thereby distinguishing between long-run effects and short-run effects.

The 2007-09 economic and financial crisis, including the run-up to this crisis, is a good example of the interactions in the above model. It has clearly shed light on the importance of the relationship between the development of debt and the development of house prices, and subsequently housing wealth. From an analytical point of view, it is worth separating high debt from low debt busts. Research done by the IMF (2012) suggest that housing busts preceded by significant increases of household debt tend to be followed by more severe and longer-lasting declines in household final consumption, and, as a consequence, a fall in GDP and an increase in unemployment. In this respect, one can raise the question whether the larger decline in household spending simply reflects the larger declines in house prices, but the empirical evidence shows that the significant drop in house prices explains at most one quarter of the stronger decline in household consumption. On the contrary, in combination with the pre-bust increase in debt, household deleveraging, i.e. the wish to run down debt levels, mainly explains the severity of the contraction of household consumption. Moreover, household deleveraging tends to be more pronounced following busts preceded by a larger growth in household debt. Overall, the IMF study concluded that when households accumulate more debt during a boom, the subsequent bust gives rise to a more severe contraction in economic activity.

The above developments are also relevant for the counterpart sectors of household debt. The impact of non-performing loans on the asset side of the creditors, mainly banks, may create spill-over effects to the creditors’ liability side. Banks are required to stabilise their own capital reserves, which indirectly also affects the owners of the banks. Based on the approach pursued by Kavonius and Carsten (2009) on the balance sheet interlinkages using FWTW data, Austria has tested equity losses due to non-performing loans of households. The hypothetical transmission of balance sheet shocks in the first ten rounds was simulated assuming a 10% loss on banks’ portfolio of loans granted to households. The results for Austria showed that the total losses in the value of equity issued by banks would amount to roughly 16% (Andreasch, 2014). In addition, banks may be affected by the presence of uncertainty among households, which can lead to precautionary saving. Although there is no direct link between saving as a resource of funds, and financial investments as a use of these funds, precautionary saving may increase financial investment in the form of deposits held with banks and consequently influences the role of households as providers of funds to the banking system.

Box 4.2. Some indicators from the household sector’s financial accounts and balance sheets, used by international organisations

International organisations use a variety of indicators that can be derived from the financial accounts and balance sheets of households, to make comparisons across countries. In this box, a limited overview is given of such indicator sets.

Households’ Net Financial Wealth Ratio (used by the OECD, as part of their Household Dashboard): The ratio of financial assets less total debt, as a percentage of Gross Household Disposable Income. When assessing vulnerabilities, one should not only look at gross debt levels, but also at the availability of assets, preferably taking into account both financial and non-financial assets (for households, predominately real estate). One should also be aware of the fact that assets are typically more prone to market price changes than liabilities. For example, for individual households, changes in house prices may easily result in mortgage debt levels which are higher than the value of the underlying dwellings, as a consequence of which the relevant households may end up in a very vulnerable financial situation.

Households’ Indebtedness Ratio (used by the OECD, as part of their Household Dashboard) and Gross Debt-to-Income Ratio of Households (Eurostat Key Indicators): The ratio of total debt (usually determined by loans) to Gross Household Disposable Income. This ratio is often used as an indicator for (changes in) financial vulnerabilities of households; it can also be used to assess debt sustainability.

Household Debt-to-GDP Ratio (used by the IMF, as part of their Financial Soundness Indicators (FSI)): The ratio of total household debt (usually determined by loans) to Gross Domestic Product (GDP).

Household Debt Service Ratio (used by the BIS): The ratio of interest payments plus amortisation of debt to Gross Household Disposable Income. The debt service ratio provides an indicator for the part of income which needs to be spent on debt and which thus cannot be used for final consumption.

Household Sector Report (used by the ECB): The ECB publishes, on a quarterly basis, the Household Sector Report for the Euro Area and for all EU member states. The report contains the following indicators from the financial accounts data:

Household financing and investment per capita
Household investments in financial assets and contribution by components
Household saving, non-financial investment, financing and debt
Changes to household net worth (financial and non-financial assets minus debt) broken down by flows and other changes in stocks.

In view of the above, the government may intervene to improve the macro-economic situation and decrease vulnerabilities in the financial system by providing temporary fiscal stimulus, automatic support to households through the social safety net, support for household debt restructuring, or assistance to the financial sector. All these interventions have an effect on net lending/net borrowing and will lead, ceteris paribus, to higher government debt, as seen during and after the 2007-09 economic and financial crisis in a number of countries.

3. Distributional aspects of wealth and going beyond households’ economic well-being

Linking macro-economic statistics and micro-data on distribution

The data derived from the financial accounts and balance sheets provide powerful information on the financial behaviour of households as a whole, but they do not provide any information on distributional issues. How large is the inequality in the distribution of wealth? Which households actually invest in listed shares during stock exchange booms? Which households ask for loans to invest in real estate? Which households save money for retirement? These are very pertinent questions, not only for economic reasons but also from a well-being perspective. In recent years, more and more work has been done to arrive at distributional information that is consistent with the data from national accounts, by trying to bridge the differences between the macro data from national accounts, and more detailed individual data from surveys (and administrative data) on income, consumption and wealth.

It is important to note that macro and micro wealth data are designed to answer different types of questions about saving and wealth accumulation. Estimates of key macro-economic parameters, such as the marginal effect of income and wealth shocks on consumption (and saving) require high frequency, timely and comprehensive macro data. On the other hand, analyses of individual, more heterogeneous saving and wealth accumulation behaviour due to differences in income, age, personal characteristics and other variables can only be derived from micro-data.

Analysing macro-economic developments in conjunction with the distribution of income and wealth across households shows that an increase of inequality in wealth often occurs in tandem with the increase in income inequality. This is due to the fact that an increase in income for groups at the top of the income distribution results in higher saving, which leads to a further accumulation of wealth. Furthermore, the composition of household balance sheets varies considerably, not only between countries but also between the lower and the upper ends of the wealth distribution. Data for France, Germany, Italy, Spain, the United Kingdom and the United States,9 for example, highlight that households in the higher wealth quintiles have a higher proportion of equity shares than households in the lower quintiles, while the distribution of real estate across quintiles also differs quite significantly. In the observed countries, these two wealth components also show to be the main drivers behind differentials in the return on assets.

Preferably, one would like to link macro- and micro data on income, consumption and wealth, to arrive at a better understanding of financial behaviour of households, and this behaviour’s impact on economic developments. However, the aggregation of individual wealth measures does not sum up to the macro data in national accounts. The reasons for these differences can be split into four groups:

differences in the definition of households;
differences in the definition of wealth;
differences in the valuation of individual portfolio items; and
the coverage of the estimates.

Regarding the definition of households, the 2008 SNA defines households as all individuals living in a certain country regardless whether they stay at home, are homeless, or are part of the so called institutionalised population (people hospitalised for longer periods, people living in military institutions, or prisoners). Furthermore, all individuals living in the country for at least one year, irrespective of periods spent abroad of less than one year, are counted as part of the household sector. In contrast, surveys capturing income, consumption and wealth of individuals select a certain sample of families representing the population, usually excluding the homeless and institutionalised population. Empirical data for the Euro Area show that the difference between these two population groupings can amount to up to 5%. Moreover, mainly due to data limitations in the compilation of national accounts, the household sector may include Non-Profit Institutions Serving Households (NPISHs), like churches, trade unions or foundations.

When it comes to the definition of wealth, the classifications and concepts of the various income and wealth items in national accounts are somewhat different from those used in household surveys. Conceptually, micro surveys focus on individual households, and therefore the definition is based on the household point of view, while in national accounts the items are defined at the level of the total economy with a subsequent split into main sectors. One of the examples of this type of divergence – at least in Europe – relates to the definition of business wealth of households managing unincorporated enterprises. In micro surveys, often only the net asset value of the enterprise, i.e. the balance of all financial and non-financial assets less liabilities, is recorded, whereas in national accounts the individual assets and liabilities of the unincorporated enterprise are classified in the relevant categories. In addition, some categories of wealth may not be available in survey based wealth measures, like cash (due to the impossibility of raising the question in an interview) or non-life insurance technical reserves (due to the lack of knowledge of households).

Looking at the valuation of individual portfolio items, one of the differences is related to the valuation of entitlements for retirement. While in micro surveys calculations of the net present value of life insurance contracts and some types of pension plans are made on the basis of premiums and length of the contract, in national accounts the relevant amounts are simply derived from the very detailed estimations of the relevant entitlements by insurance corporations and pension funds. In addition, the valuation of other individual financial instruments may differ. The valuation required by the 2008 SNA follows the market valuation principle as closely as possible. In practice, this is done through the use of centralised information or counterpart information. For example, the valuation of tradeable securities held by households in financial accounts and balance sheets follows the concept of market quotations. Likewise, the main portion of household debt is held as assets by banks and is valued accordingly. Furthermore, in the absence of relevant market price information, equity holdings in limited liability companies are usually valued according to the value of the own funds or book value. In contrast, households may not be in the position to value the assets with the same standards as used for national accounts.

A final point relates to the coverage of the estimates, that is, whether the estimates provide a full and complete picture of the transactions or positions under consideration. The explicit goal of national accounts is to arrive at exhaustive estimates, including hidden and illegal activities. Moreover, the balancing framework of national accounts aims at reaching consistency between all types of assets and the corresponding liabilities for each financial instrument, thus using – when considered appropriate – more reliable counterparty information which tends to lead to a better coverage of the estimates. Moreover, efforts to try to arrive at consistency between non-financial transactions, financial transactions and (changes in) positions on the balance sheets lead to a further improvement of coverage and quality more generally. These considerations do not generally play a role in the survey based compilation of wealth of households. Moreover, the sampling of households usually leads to an underestimation of wealth, even taking into consideration oversampling of certain groups for the improvement of the response rate. It is, for example, of crucial importance to have a good representation of the top 1% of wealthy households, as they have a massive impact on the estimates. However, this would involve the need to include a proper representation in the survey, which is very difficult.

All in all, a naïve comparison, without any adjustments for differences in concepts and methodologies, leads to a lower coverage of financial wealth in survey data, if compared to data derived from financial accounts and balance sheets. These results call for a compilation based on a method in which the definitions and methodologies are aligned as much as possible. Such a comparison is available for the United States for a number of years, showing that survey data actually reports higher figures than the macro data derived from national accounts, mainly in the area of non-financial assets, since 2010. Provisional results show that, after aligning the micro and macro data to the same concepts and methodologies, the results from micro data, as a percentage of macro data, would increase between 10 and 20 percentage points for the majority of Euro Area countries, the ratio of micro data thus amounting to between 40% and 60% of the adjusted financial accounts data.

Beyond financial accounts and balance sheets: households’ well-being

The standard national accounts, which are compiled by institutional sector, generate a broad range of indicators for economic well-being, such as household (disposable) income, saving and (financial) wealth. This has also been acknowledged by the so-called “Stiglitz-Sen-Fitoussi Report”10 by the Commission on the Measurement of Economic Performance and Social Progress. One of the Commission’s goals was “… to identify the limits of GDP as an indicator of economic performance and social progress ... [and] to assess the feasibility of alternative measurement tools”. In its recommendations, the report concludes that, when evaluating material well-being, one should i) look at income and consumption rather than production; ii) emphasise the household perspective; and iii) consider income and consumption jointly with wealth. Other suggestions include iv) to give more prominence to the distribution of income, consumption and wealth; and v) to broaden income measures to non-market activities, such as unpaid housework.

Put differently, these recommendations encourage broadening and deepening the analysis of household behaviour, as noted in the conclusion of the report “… that in the measurement of household welfare all material components should be covered, i.e. consumption, income and wealth, from both the micro as well as the macro perspective”. The financial accounts and balance sheets cover the accumulation of household financial wealth and the resulting stocks. Measures of the economic relevance of households include the contribution of household consumption to GDP growth and their share of net lending in the economy, providing funds both as investors and savers to other domestic and external economic sectors.

The measurement of a more encompassing concept of well-being is high on the agenda of the statistical community. The OECD plays a key role in this respect, amid initiatives of the United Nations Economic Commission for Europe (UNECE), the European Commission, and similar initiatives by several advanced and emerging countries. The framework of well-being focuses on the measurement of broader metrics representing the development of well-being of people in each country, in addition to the macro-economic conditions using standard macro-economic performance indicators like GDP. The main goal is to identify the drivers of well-being. Household debt and household (financial) worth are part of this broader set of indicators of well-being.

The indicators used in the context of the broader framework also include material living conditions such as income and wealth, jobs and earnings, and housing. The economic well-being, in the sense of people’s command over economic resources, is a multi-dimensional concept whose components (income, consumption and wealth) are separate but interrelated. Looking at these different types of economic resources jointly allows for an improved identification of people in distressed or (dis)advantaged conditions, thus also allowing a better targeting of policies. However, there are numerous other aspects affecting the well-being of people, such as health status, work and life balance, education and skills, social connections, civic engagement and governance, environmental quality, personal security and subjective well-being indicators; see Figure 4.15. A concrete example of such a broader set of indicators trying to capture well-being is the OECD Better Life Index.11 A dashboard providing a set of more economically oriented indicators can be found in the OECD Dashboard on Households’ Economic Well-being.

Summarising recent developments, it is evident that understanding financial and non-financial behaviour of households has become more important. This leads to a demand for better and more granular statistics on households, which financial accounts and balance sheets can partially help to provide.

Key points

Households act both as an important driver of economic growth, notably in the form of final consumption expenditure, and as a key player in providing funds to other sectors. They also act as producers of goods and services by operating small and medium-serviced enterprises, of which some are incorporated.
Households in OECD countries partially converged the structure of their asset portfolio, and there is a growing importance in most countries for long-term investments related to pension plans.
Non-financial wealth, especially in the form of dwellings, has become more important. Often this investment in non-financial wealth is accompanied with a notable increase of household debt, which is now the subject of macro-economic stability analysis, especially after the bubble in house prices and the resulting economic and financial crisis in 2007-09.
Households in national accounts are grouped together into one “institutional sector”, as a consequence of which monitoring and analysis is only possible at the aggregate level. In recent years, there is a growing interest for more detailed analysis for certain subgroupings of households, for which it is necessary to elaborate micro data statistics, trying to align these data to the aggregates from national accounts.
More generally, it has to be noted that economic indicators only tell a partial story when it comes to the well-being of people. Broader measures have been developed and continue to be improved to capture various (and beyond the economic) elements affecting people's well-being.

Going further

Captured “residually” – Data sources for households in financial accounts and balance sheets

Data on households’ financial investments and related financial assets, as well as external financing and related debt, are compiled from a variety of data sources, including balancing procedures using counterparty information. As a consequence, the data are often compiled on a residual basis (often including NPISHs). Table 4.2 gives some indication of the methodologies typically used, but it is important to mention that there are no general rules due to the heterogeneity of data sources and compilation procedures across countries.

Table 4.2. Summary of data sources for financial instruments
Financial instrument	Definition	Domestic assets/liabilities	Foreign assets/liabilities
Financial assets
Currency	Banknotes and coins including silver and gold coins (provided that they can be exchanged for cash at any time). Gold bullions are not included.	Residual item, usually derived from liability side minus holding by domestic financial sector and estimated value for cross-border holdings	Balance of payments statistics (BOP), estimates
Deposits	Current or saving accounts (including saving cards) with domestic and foreign banks, regardless of maturity	Counterpart data from banks	Direct reporting or counterpart statistics from non-resident banks
Debt securities, listed shares, investment fund shares	Debt securities are commercial papers, short-term treasuries, and all kind of bonds (zero-bonds, covered bonds, etc.). Listed shares and investment fund shares are shares which are officially listed on a stock exchange, or traded in any other form of secondary market	Custodian reports or residual compilation based on issuance statistics, statistics on holdings of the financial sector and the government sector, estimates on the split between the holdings of non-financial corporations and those of households, survey data	Direct reporting or transaction-based reporting by banks for BOP statistics
Unlisted shares, other equity	Unlisted shares, or equity issued by limited liability companies	Company register, direct reporting, custodian reporting, residual compilation (see listed shares), survey data	Direct reporting, data derived from Foreign Direct Investment (FDI) statistics, transaction-based reporting by banks for BOP statistics
Life insurance contracts	Actuarial reserves and other technical reserves for entitlements related to individual non-life and life insurance policies	Counterpart data from insurance corporations	Direct reporting, transaction-based reporting by banks for BOP statistics (often based on premiums)
Pension entitlements	Entitlements related to private (funded) pension schemes, provided directly by employers for their employees, or more usually via separated pension funds. Payasyougo systems are not included	Counterpart data from pension funds, banks, insurance corporations and company balance sheet data	Direct reporting, transaction-based reporting by banks for BOP statistics (often based on premiums)
Financial derivatives	Employee stock options offered as part of salaries, tradeable derivatives like options	Custodian reports, direct reporting, data on employee stock options derived from non-financial accounts data, survey data	Direct reporting, transaction-based reporting by banks for BOP statistics
Loans, trade credits	Loans between households, and loans owed by companies with an equity stake of households; trade credits in the form of pre-paid vouchers	Company balance sheet data, survey data	Direct reporting, transaction-based reporting by banks for BOP statistics
Other financial assets	Tax credit, accrued interest on deposits, if not part of the related financial instrument, money in transit	Dependent on national circumstances	Dependent on national circumstances
Liabilities (Debt)
Loans	Housing loans, consumer loans, and other loans (mainly consisting of loans extended to self-employed individuals and sole proprietorships for investment purposes; loans for debt consolidation, education and retirement investment; and salary advances)	Counterpart data from banks, other financial institutions, government	Direct reporting or counterpart statistics from non-resident banks
Trade credits	Trade credits (difference between delivery and payment) of suppliers for goods and services	Company balance sheet data	BOP-statistics
Pension entitlements	Additional payment liabilities for funded pension schemes by the employees	Counterpart data from pension funds	BOP-statistics
Other	Tax credit, accrued interest on loans, if not part of the related financial instrument loans, money in transit	Dependent on national circumstances	Dependent on national circumstances

References

Andreasch, M. (2014), “Analysis of the financial interlinkages of the financial sector in Austria against the background of the recent financial crises”, in A Flow-of-Funds Perspective on the Financial Crisis, Vol. II, Palgrave Macmillan Studies in Economics and Banking, Basingstoke.

Andreasch, M. et al. (2010), “Investment and financing activities of the institutional sectors of the Austrian economy 2010”in Sector Accounts in Austria 2010, Oesterreichische Nationalbank, Vienna, https://www.oenb.at/dam/jcr.../shst_2010_ june_sector_accounts_tcm16-236095.pdf.

Antoniewicz, R et al. (2005), “Household wealth: Comparing micro and macro data in Cyprus, Canada, Italy and United States”, paper prepared for the LWS Construction and Usage of Comparable Microdata on Wealth: The LWS Workshop, Bank of Italy, Perugia, 27-29 January 2005.

BIS (2017), Debt Service Ratios, Bank for International Settlements, Basel, www.bis.org/statistics/dsr.htm.

Bruno, G. and R. De Bonis (2008), “Do financial systems converge? New evidence from household financial assets in selected OECD countries”, OECD Statistics Working Papers, No. 2009/1, OECD Publishing, Paris, https://doi.org/10.1787/224175173554.

Bouis, R. (2015), “Household deleveraging and saving rates: a cross-country analysis”, paper prepared for the KDI International Conference on Household Debt from an International Perspective: Issues and Policy Directions, Korea Development Institute, Seoul, 10 July 2015.

Coletta, M., R. De Bonis and S. Piermattei (2014), “The determinants of household debt: A cross-country analysis”, Economic Working Papers, No. 989, Bank of Italy, Rome, www.bancaditalia.it/pubblicazioni/temi-discussione/2014/2014-0989/en_tema_989.pdf.

De Bonis, R., D. Fano and T. Sbano (2012), “Household wealth in the main OECD countries from 1980-2011: What do the data tell us?”, Questioni di Economia e Finanza (Occasional Papers), No. 160, Bank of Italy, Rome, www.bancaditalia.it/pubblicazioni/qef/2013-0160/QEF_160.pdf.

Domanski, D., M. Scatigna and A. Zabai (2016), “Wealth inequality and monetary policy”, BIS Quarterly Review, March, Bank for International Settlements, Basel, www.bis.org/publ/qtrpdf/r_qt1603f.htm.

Eurostat (2016), Sector Accounts: Households and NPISH, Eurostat, Luxembourg, http://ec.europa.eu/eurostat/web/sector-accounts/detailed-charts/households-npish.

Honkkila, J. and I. Kavonius (2013), “Micro and macro analysis on the household income, wealth and saving in the Euro Area”, ECB Working Paper Series, No. 1619/November 2013, ECB, Frankfurt, www.ecb.europa.eu/pub/pdf/scpwps/ecbwp1619.pdf.

IMF (2012), “Dealing with household debt” in World Economic Outlook April 2012, IMF, Washington, DC, www.imf.org/external/pubs/ft/weo/2012/01/pdf/c3.pdf.

Kavonius, L. and O. Carsten (2009), “Balance sheet interlinkage and macro-financial risk analysis in the Euro Area”, ECB Working Paper Series, No. 1124/December 2009, ECB, Frankfurt, www.ecb.europa.eu/pub/pdf/scpwps/ecbwp1124.pdf?b1f1ab145ac34f 5c0ed908641d1e2f9d.

Magri, S. (2009), “Household wealth and entrepreneurship: Is there a link?”, Economic Working Papers, No. 719, Bank of Italy, Rome, www.bancaditalia.it/pubblicazioni/temi-discussione/2009/2009-0719/en_tema_719.pdf.

OECD (2017), “Aggregate National Accounts, SNA 2008 (or SNA 1993): Gross domestic product”, OECD National Accounts Statistics (database), https://doi.org/10.1787/data-00001-en; and “Detailed National Accounts, SNA 2008 (or SNA 1993): Non-financial accounts by sectors, annual”, OECD National Accounts Statistics (database), https://doi.org/10.1787/data-00034-en; and “Financial Accounts, SNA 2008 (or SNA 1993): Consolidated flows, annual”, OECD National Accounts Statistics (database), https://doi.org/10.1787/data-00716-en; and “Financial Balance Sheets, SNA 2008 (or SNA 1993): Non-consolidated stocks, annual”, OECD National Accounts Statistics (database), https://doi.org/10.1787/data-00720-en.

OECD (2014), Understanding National Accounts, OECD Publishing, Paris, https://doi.org/10.1787/9789264214637-en.

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

OeNB (2017), National Accounts, Osterreichische Nationalbank, Vienna, www.oenb.at/en/Statistics/Standardized-Tables/financial-accounts.html.

Turk, R. (2015), “Housing price and household debt interactions in Sweden”, IMF Working Papers, No. 15/276, IMF, Washington, DC, www.imf.org/external/pubs/ft/wp/2015/wp15276.pdf.

Van der Wal, D. (2014), “The measurement of international pension obligations – Have we harmonised enough?”, DNB Working Papers, No. 424/May 2014, De Nederlandsche Bank, Amsterdam, www.dnb.nl/binaries/Working%20Paper%20424_tcm46-307802.pdf.

Notes

← 1. 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.

← 2. These countries comprise Austria, Belgium, Czech Republic, France, Germany, Hungary, Italy, Japan, Korea, Mexico, Portugal, Slovenia, Sweden and Switzerland.

← 3. Here, gross disposable income refers to “adjusted” disposable income. It thus includes social transfers in kind, i.e. income equivalent to goods and services provided by government for free or at prices that are not economically significant (such as health and education).

← 4. These countries comprise Estonia, Ireland, Netherlands, Norway, Poland, Slovak Republic, Spain, United Kingdom, and the United States.

← 5. The recording of pension schemes slightly complicates this calculation. It is necessary to not use disposable income as such, but to add the adjustment for the change in pension entitlements. It is this total figure that should be the denominator in the saving ratio calculation. See Chapter 9 for further details.

← 6. For more information, see www.oecd.org/std/na/Measuring-inequality-in-income-and-consumption-in-a-national-accounts-framework.pdf.

← 7. See www.oecd.org/std/na/household-dashboard.htm.

← 8. For more details see www.bis.org/statistics/dsr/dsr_doc.pdf.

← 9. Data are derived from the Eurosystem Household Finance and Consumption Survey, data provided by the UK Office of National Statics, Wealth and Assets, and the Survey of Consumer Finances of the US Federal Reserve Board.

← 10. The Stiglitz-Sen-Fitoussi Report is the closing report of the “Commission on the Measurement of Economic Performance and Social Progress”. This commission was set up, at the start of 2008, at the initiative of the French government. The full report is available at the following link: http://ec.europa.eu/eurostat/documents/118025/118123/Fitoussi+Commission+report.

← 11. See www.oecdbetterlifeindex.org/.