How Rich Are You?

How rich are you? Intuitively, this is a simple question, which everyone should be able to answer with a single numerical answer. However, as Asli discussed in last month’s blog, wealth is a complex concept, and the measurement of it is far from straightforward.

As a part of the WEALTHPOL project, I have looked over numerous surveys and datasets which aim at estimating the wealth owned by individuals or households. One thing which I found striking is the variety in approaches to wealth measurement across surveys. The most straightforward surveys ask only one question, such as "[Enter] total household assets" (from the National Social Life, Health, and Aging Project) or "Please estimate your total wealth" (from the General Social Survey). On the other hand, the most complex surveys have more than 250 questions related to wealth. Two of the most detailed surveys are the Wealth and Asset Survey and the Luxembourg Wealth Study. These latter cover everything from the value of the interviewee’s house to the value of gifts they have received.

Naturally, the way we ask the question matters for the quality of the data and for the type of analyses we can do subsequently. Researchers must therefore think hard about how to design their surveys. For example, some surveys are known to systematically lead to the underreporting of wealth.[1]

It may be useful here to go back to the basic definition of wealth. In the Oxford English Dictionary wealth is defined as 'an abundance of valuable possessions or money'. Wealth can therefore be perceived as income that we do not spend, which is therefore free to be invested (or not). The invested money becomes a cumulative stock that reflects years of prior circumstances and decisions. When we measure wealth, we should try to measure this stock at a specific point in time. Practically, we typically measure wealth as net worth. This is defined as the sum of the value of a household’s assets, less the value of debts.[2]

One survey which does the best job in measuring wealth is The Panel Study of Income Dynamics (PSID). PSID differentiates between the following wealth related items:

1. Housing wealth: Does the respondent own a house? What is the current value of the house?
2. Other real estate: What is the value of any other real estate (other than the main home) such as a second home, land, rental real estate, or money owed on a land contract?
3. Vehicle wealth: What is the value owned on wheels? This can be cars, trucks, a motor home, motorcycles or boats.
4. Farm or Business: Does the respondent own a part or all of a farm or business? What is the value of this?
5. Stocks, bonds and Treasury Bills: What is the value of shares of stock in publicly held corporations, treasury bills, stock mutual funds, or investment trusts, not including stocks in retirement accounts?
6. Annuities and Individual Retirement Account (IRA): What is the value of money in private annuities and Individual Retirement Accounts?
7. Checking, Savings, Money Market Accounts: What is the value of money in checking or savings accounts, including money market accounts?
8. Other wealth: What is the value of other savings or assets, such as bond funds, cash value in a life insurance policy, a valuable collection for investment purposes, furniture, patents, or rights in a trust or estate?
9. Debts: Does the respondent have any debts? This can be mortgage on the main home, credit card charges, student loans, medical or legal bills, or loans from relatives.[3]

These questions should cover all types of investment decisions, and give a comprehensive picture of personal wealth. However, the approach used in PSID may not work in every survey setting. It is demanding on the respondent, requires many questions, and may generate data too detailed to be useful for analytical purposes. An alternative approach is to ask one question such as "Please estimate your total wealth", and accompany this question with a definition such as "Wealth means the value of your house plus the value of your vehicles, stocks, and mutual funds, cash, checking accounts, retirement accounts including 401(k) and pension assets, and any other assets minus what you owe for your mortgage and your debts", as the GSS does.[4]

In addition to the phrasing of the questions themselves, other elements are worth thinking about when designing a survey about wealth. Should the answers be categorical or continuous? Do we want to measure personal or household wealth? If we want to measure household wealth, it is preferable to also ask about the size of the household. Furthermore, we may want to oversample specific groups of people such as very affluent people, who often are underrepresented in surveys.[5]

In sum, while there is no perfect way of asking How rich are you?, there certainly are many wrong ways to do it. Putting an extra effort into designing the survey is therefore essential. It determines the amount and quality of information you can extract from such surveys; and this information in turn inform the way individuals and researchers alike think about wealth.