Household debt-to-income ratios are at historically high levels in many advanced economies. Many households now carry debt close to, and well into, retirement (Lusardi et al. 2018). This high indebtedness means both households and entire economies are vulnerable to shocks, which is a risk to the sustainability of growth (OECD 2017).
Figure 1 Evolution of household debt and disposable income in a sample of advanced economies
Source: OECD (2017).
Notes: Simple average of Australia, Austria, Belgium, Canada, Czech Republic, Germany, Denmark, Estonia, Finland, Hungary, Italy, Japan, Netherlands, Norway, Slovak Republic, Spain, Sweden, Switzerland and US.
Because this household debt can have economy-wide effects, it is important that we improve our understanding of what is determining it, and why it has recently increased in many countries. Economic fundamentals, such as lower interest rates and the deregulation of financial services, are likely to have been important, but do not explain the large variation among households in a country, or among households in different countries.
Attitudes to debt
Research has indicated that cultural differences may play a role in shaping debt choices, given that apparently similar countries can differ in their credit arrangements (Badarinza et al. 2016). Differences in credit market development or tax rules are unlikely to fully explain large cross-country differences in, for example, credit card use, the pervasiveness of mortgages, or the dominance of adjustable or fixed-rate mortgages (Campbell 2013).
Social norms linked to borrowing and saving may be an important element of these cultural differences. Researchers have found these norms in many societies. This history is reflected in language, religion, and culture – in many Indo-European languages, words for debt often also mean 'sin' or 'guilt', and several religions, including Christianity and Islam, have condemned interest on loans (Graeber 2011). In many cases, governments or civil society have propagated social norms that encourage saving, for example through savings schemes or informational programs (Garon 2013).
We argue that changing attitudes may have contributed to the build-up of household debt. Attitudes to debt have been linked to specific choices, in particular credit card use (e.g. Chien and Devaney 2001). In a recent working paper (Almenberg et al. 2018), we introduce a simple and novel measure of debt attitude: whether people feel uncomfortable with debt. By investigating attitudes toward debt we can understand better whether individual debt choices are affected by social norms. These norms have been linked to individual decision making in many areas, including wage-setting (Akerlof 1980) and consumption (Elster 1989).
Explaining levels of personal debt
The data we use are from Sweden, where household debt in relation to GDP has nearly doubled in two decades to about 90% in 2018. The survey sample is representative of the Swedish population age 25–751 and it consists of 390 men (46% of the sample) and 454 women (54%). The average age is 51. The survey was carried out by Statistics Sweden on our behalf, and matched with register data.
Our study shows that more than half of the respondents in the survey report are uncomfortable with debt. This simple indicator helps explain individual debt levels. Linking the survey data with register data, we show that those who report being uncomfortable with debt have considerably lower debt-to-income ratios. On average, the difference in debt-to-income is about 75% of annual disposable income ($25,000). Controlling for observable differences slightly reduces the difference to 60%.
Parents are an important channel for the transmission of social norms. To shed more light on this, survey respondents were also asked about their parents’ attitudes to debt. There is a strong correlation between the debt attitudes of parents and their children, suggesting intergenerational transmission of attitudes toward debt.
Figure 2 displays the share of respondents who report being uncomfortable or comfortable with having debt, broken down by their parents’ attitude to debt. The share of uncomfortable respondents is 70.4% among the group whose parents are uncomfortable with debt. In contrast, the share of respondents uncomfortable with debt is only 27.7% among the group whose parents are comfortable with debt.
Conversely, the share of comfortable respondents is 72.3% among the group whose parents are comfortable with debt but only 29.6% among the group whose parents are uncomfortable. Thus, there is a strong association between parents’ attitude toward debt and the attitude of their children.
Figure 2 Respondents’ and parents’ attitude to having debt
Source: Almenberg et al (2018).
Notes: The bar charts show the respondents’ answer to the question “Do you feel uncomfortable with having debt?”, broken down by parents’ attitude to having debt. The vertical axes indicate the shares of respondents. The whiskers indicate the 90% confidence interval.
Moreover, our survey suggests a change in attitudes over time. As shown in Figure 3, respondents in recent cohorts (baby boomers and after) are more likely to be comfortable with debt. Part of the increase in debt in recent decades may simply be due to a change in attitudes toward debt.
Figure 3 Debt attitude of respondents and their parents by year of birth
Source: Almenberg et al (2018).
Notes. The bars show the share of respondents (right panel) and respondents’ parents (left panel) that were reported to be uncomfortable with debt, by birth cohort (grouped). The horizontal axes refer to birth year of the respondents (right) and parents (left), respectively.
If you are sceptical that some people feel more uncomfortable with debt than others, ask people of different ages and different cultural backgrounds. Ask also about their parents. Your anecdotal evidence should show that there is a great deal of variation in attitudes among individuals.
Relationship to existing theory
This is not a rejection of the standard theory of consumption and saving, but a possible enrichment of that model. A simple intertemporal model relies on consumption smoothing to generate predictions about individual decisions to save or borrow, and adding uncertainty generates a richer model.
Allowing for preference heterogeneity, variation in self-control, or varying levels of financial literacy, gives rise to richer patterns of saving and borrowing than can be explained by the standard life-cycle model alone. We have found an additional potential determinant of intertemporal choices about consumption, namely that some people may refrain from borrowing because they have been taught to do so – in other words, they have internalised a social norm that discourages debt. In terms of the standard economic theory, social norms that shape attitudes toward debt would, in effect, be acting as a constraint on individual borrowing behaviour.
If we accept that social norms that discourage borrowing exist. In general, social norms act as an additional constraint on behaviour, driving individual decisions toward conformity. Can we say that these norms are 'good' or 'bad'? As economists, we argue that it’s not a clear-cut case.
- Good: more efficient outcomes. For example, norms against cheating or free-riding may mitigate moral hazard or time inconsistency problems, resulting in more efficient outcomes. Debt can lead to financial distress if not managed properly (Lusardi and Tufano 2015). Some borrowers may underestimate the future debt burden associated with a loan, because they have limited financial literacy or exponential growth bias, or succumb to temptation and incur debts in a time-inconsistent manner. In these cases, a social norm that leads to less debt could be welfare improving.
- Good: reduces externalities. At the aggregate level, there might be negative externalities from high household debt, for example if the economy becomes less resilient to economic shocks. Individual households are unlikely to take such externalities into account in choosing their privately optimal debt level. In this case, a norm against debt could generate an outcome that is closer to the socially optimal debt level.
- Bad: restricts some markets. Social norms that label transactions of certain goods as repugnant can shut down markets entirely (Roth 2007). A social norm against debt that deters people from investing in human capital is socially suboptimal, if there are positive externalities from such investment.
- Bad: reduces consumption smoothing. A social norm against debt may inhibit smoothing in a way that is similar to a liquidity constraint, with comparable welfare losses.
- Bad:cost of capital. Social norms against debt could also explain why many households pay a sizable premium for less salient debt, raising their cost of capital (Almenberg and Karapetyan 2014).
In sum, social norms related to debt could both increase and decrease economic efficiency. Our analysis does not attempt to assess which of these views carries more weight. But we believe that norms deserve more attention than we have given them so far in economic models.
Almenberg, J and A Karapetyan (2014), “Hidden Costs of Hidden Debt”, Review of Finance 18(6): 2247-2281.
Almenberg, J, A Lusardi, J Säve-Söderbergh and R Vestman (2018), “Attitudes to Debt and Debt Behavior”, NBER working paper 24935.
Badarinza, C, J Y Campbell and T Ramadorai (2016), “International Comparative Household finance”, Annual Review of Economics 8(1): 111-144.
Campbell J Y (2013), “Mortgage Market Design”, Review of Finance 17(1): 1–33.
Chien, Y and S A Devaney (2001), “The Effects of Credit Attitude and Socioeconomic Factors on Credit Card and Installment Debt”, Journal of Consumer Affairs 35(1): 162–179.
Garon, S (2013), Beyond Our Means: Why America Spends While the World Saves, Princeton University Press.
Graeber, D (2011), Debt: The first 5,000 years, Melville House.
Lusardi, A, O S Mitchell, and N Oggero (2018), “The Changing Face of Debt and Financial Fragility at Older Ages”, American Economic Association Papers and Proceedings 108: 407–411.
Lusardi, A, and P Tufano (2015), “Debt Literacy, Financial Experiences, and Overindebtedness”, Journal of Pension Economics and Finance14(4): 329–365.
OECD (2017), OECD Economic Outlook 2017, OECD Publishing.
Roth, A E (2007), “Repugnance as a Constraint on Markets”, Journal of Economic Perspectives 21(3): 37–58.
 The sample was generated using the registry for the total population, which contains 6.1 million individuals in the chosen age span. A total of 2,004 individuals were drawn from ten strata based on age and gender. Thirty-five of these individuals were excluded (due to incarceration, etc.), resulting in a sample of 1,969 individuals. For each of these individuals, at least twelve attempts to establish contact were made during eight weeks between September and November 2014. After this time period, 844 individuals had responded. Out of these, the total number of individuals with non-missing values on our debt measures, attitudes and control variables is 727 individuals.