VoxEU Column Financial Markets

The causes of mortgage default: Ability to pay versus negative equity

Many studies have addressed the question of why people default on their mortgages, but lack of data has meant that much of this research has omitted the effect of the owner's ability to pay. This column uses panel data on defaults and changes in income to show that ability to pay is a much more important determinant of default than previously recognised. If the head of household loses a job, for example, this is equivalent to the effect of a 35% drop in home equity. Policies targeted at increasing ability to pay may be more effective at reducing default than those that try to remedy negative equity.

Why do homeowners default on their mortgage? Goodman et al. (2010) suggest two possible forces. One is a lack of liquidity – homeowners no longer have the ability to pay their mortgage because they have suffered a significant negative income or expenditure shock. The other is negative equity, often referred to as 'strategic default'. In this case homeowners have the ability to pay but choose to default because they have high negative equity.

Despite a lot of research on the determinants of mortgage default, we don't know for sure the relative importance of ability to pay and strategic motivations. Lack of data has meant that much of the research has omitted the effect of the owner's ability to pay. Instead, it focuses on the role of negative equity (Vandell 1995, Deng et al. 2000). It finds that negative equity is the main driver of defaults, but at the same time acknowledges that factors such as liquidity or ability to pay are potentially important. Foote et al. (2008) found that equity alone was not a very accurate predictor of default, and Gyourko and Tracy (2013) presented evidence that estimated default probabilities based only on negative equity may be biased because the calculations omit ability-to-pay variables.

It is hard to disentangle the contributions of ability to pay and negative equity to decisions to default, because default waves tend to occur around periods of macroeconomic weakness, such as the financial crisis of 2007-2009 and its aftermath. This means that negative equity, and factors that negatively affect the ability to pay such as job loss, are likely to be highly correlated. If we don't know anything about the owners' ability to pay their mortgages, it is difficult even to identify default. For example, if a defaulter is both unemployed and has negative equity, is that homeowner defaulting for strategic or ability-to-pay reasons?

Confronting these issues requires matched data that allow researchers to measure the borrower's ability to pay as well as mortgage status. Most previous studies have been conducted using only mortgage default and loan characteristic data. These data did not provide information to allow researchers to measure homeowners’ ability to pay, such as household income, employment status, assets, or non-mortgage liabilities.

New measures of ability to pay

In our recent research (Gerardi et al. 2016), we construct the first dataset that matches detailed household economic and demographic information with homeowner mortgage characteristics data. The economic information is taken from the Panel Study of Income Dynamics (PSID), matched with the PSID’s Mortgage Distress Survey, carried out between 2009 and 2013.

This allows us to measure homeowner ability to pay in much greater detail, and much more accurately, than in previous studies. We use this to document the ability to pay and equity status of both defaulting and non-defaulting households, and estimate the marginal effect of changes in ability to pay, and changes in equity, on the probability of defaulting. Measuring ability to pay means we can develop a new methodology for identifying strategic default that is based much more on economic theory than in previous research.

To implement our method, we exploit the newly matched data from the 2009-2013 PSID, including default information linked to the key components of a mortgagor’s budget constraint, including income, employment, consumption, and assets. We have not found other studies of mortgage default that use information on job loss (or any other contemporaneous factors) that significantly affect budget constraints.

Strategic default versus ability-to-pay default

Our dataset allows us to classify defaulters in terms of those who were able to pay, and those who were not. To measure strategic default, we directly measure the mortgagors’ ability to pay using their income, less obligated mortgage payments. We classify mortgagors as strategic defaulters if they could have afforded their typical consumption expenditures after paying their mortgage, at the time of default.

To assess this ability to pay, we begin with the simplest measure of total financial resources: monthly income adjusted for head and spousal job loss, divorce, and other contemporaneous income shocks. We then deduct the individuals’ monthly mortgage payment from their monthly income. We define income less mortgage payment as residual income, a concept that is widely used by mortgage originators.

Next, we isolate mortgage defaulters in all available years (2009, 2011, and 2013) and measure their consumption in the year before their default. We then classify defaulters as strategic if they could have afforded the previous year’s consumption from residual income. We find that 38% of defaulters were strategic by this measure.

The marginal effect on the decision to default

To quantify the relative importance of changes in equity and ability to pay, we specify empirical equations that related each homeowner’s decision to default or pay to variables that affect ability to pay and negative equity, as well as to demographic factors.

There are empirical challenges when estimating the causal effects of income shocks and negative equity, so we use instrumental variables. To instrument for equity, we exploit the long panel aspect of the PSID, and we use state-level house price growth from the original date of purchase as a proxy for equity. The assumption was that state-level house prices trends are strongly related to an individual’s equity position, and state-level house price growth only impacts the default decision through its effect on the homeowner’s equity position. To instrument for job loss, we use two approaches. One uses involuntary job loss as a measure of exogenous job loss, such as plant closures and lock-outs; the other uses severe, work-impairing disability. We then generate income losses from these exogenous events to use in our mortgage default analysis.

The instrumental variable results allowed us to compare the relative importance of ability to pay and negative equity. They also allowed us to look again at research that predicts mortgage default as a function of mortgage characteristics (see Bhutta et al. 2010). We find that, if the head of household loses a job this has an impact on default that is equivalent to the effect of a 35% drop in home equity. If both the household head and the spouse lost jobs, the effect on defaulting is equivalent to roughly a 50% drop in home equity.

These results indicate that that the ability to pay a mortgage is a critical determinant of default, and that defaults are likely to rise significantly during periods of macroeconomic weakness.

Policy implications

Our findings provide an answer to why lenders do not willingly renegotiate loans, even for the most at-risk borrowers. Specifically, our results show that about 90% of borrowers who have very limited ability to pay chose to continue to make mortgage payments. Consequently, low default rates among borrowers with the lowest ability to pay complicate loss mitigation policies, since the size of a payment or principal reduction that a lender is willing to offer to a distressed homeowner is increasing in the probability of that borrower defaulting. Thus, low default probabilities among distressed borrowers reduce the incentives of lenders to renegotiate loans ex-ante. Furthermore, we find that more than one-third of defaulters would need a full 100% payment reduction in order to make the mortgage affordable. This mitigates lender incentives to write down mortgages ex-post.

These results indicate that mortgagor’s ability to pay is a much more important determinant of default than previously recognised, and that loss mitigation policies targeted at increasing ability to pay, such as monthly payment modifications, may be more effective at reducing default than policies that try to remedy negative equity, such as principal reduction.


Bhutta, N, H Shan, and J Dokko (2010). "The depth of negative equity and mortgage default decisions."

Deng, Y, J M Quigley, and R Order (2000). "Mortgage terminations, heterogeneity and the exercise of mortgage options." Econometrica 68(2): 275-307.

Elul, R, N S Souleles, S Chomsisengphet, D Glennon, and R M Hunt (2010). "What 'Triggers' Mortgage Default?."

Foote, C L, K Gerardi, and P S Willen(2008). "Negative equity and foreclosure: Theory and evidence." Journal of Urban Economics 64(2): 234-245.

Gerardi, K, K F Herkenhoff, L E Ohanian, and P Willen (2016). "Unemployment, negative equity, and strategic default." 

Gyourko, J, and J Tracy(2014). "Reconciling theory and empirics on the role of unemployment in mortgage default." Journal of Urban Economics 80: 87-96.

Goodman, L S, R Ashworth, B Landy, and K Yin (2010). "Negative equity trumps unemployment in predicting defaults." The Journal of Fixed Income 19(4): 67.

Guiso, L, P Sapienza, and L Zingales (2013). "The determinants of attitudes toward strategic default on mortgages." The Journal of Finance 68(4): 1473-1515.

Vandell, K D (1995). "How ruthless is mortgage default? A review and synthesis of the evidence." Journal of Housing Research 6(2): 245.

5,675 Reads