The Wealthy-Hand-to-Mouth | Kaplan, Violante, Weidner

Greg Kaplan, Giovanni Violante, Justin Weidner; The Wealthy-Hand-to-Mouth; Brookings Papers on Economic Activity; Spring 2014 Conference; Brookings Institute; 2014-03-20; 65 pages; landing.

Abstract

The wealthy hand-to-mouth are households who hold little or no liquid wealth (cash, checking, and savings accounts), despite owning sizable amounts of illiquid assets (assets that carry a transaction cost, such as housing or retirement accounts). This portfolio configuration implies that these households have a high marginal propensity to consume out of transitory income changes–a key determinant of the macroeconomic effects of fiscal policy.

The wealthy hand- to-mouth, therefore, behave in many respects like households with little or no net worth, yet they escape standard definitions and empirical measurements based on the distribution of net worth. We use survey data on household portfolios for the U.S., Canada, Australia, the U.K., Germany, France, Italy, and Spain to document the share of such households across countries, their demographic characteristics, the composition of their balance sheets, and the persistence of hand-to-mouth status over the life cycle. Using PSID data, we estimate that the wealthy hand-to-mouth have a strong consumption response to transitory income shocks. Finally, we discuss the implications of this group of consumers for macroeconomic modeling and policy analysis.

Mentions

Definition: households who hold little or no liquid wealth (cash, checking, and savings accounts), despite owning sizable amounts of illiquid assets (assets that carry a transaction cost, such as housing or retirement accounts). [from the abstract]

  • Hand-to-Mouth (W-HtM)
  • Wealthy Hand-to-Mouth (W-HtM)
  • Non Hand-to-Mouth (N-HtM)
  • Poor Hand-to-Mouth (P-HtM)

Thesis: these households have a high marginal propensity to consume out of transitory income changes–a key determinant of the macroeconomic effects of fiscal policy. [from the abstract]

Implications:

  • W-HtM are  a distinct class
    • Centered around middle age.
    • The behavioral experience is transitory.
    • They cure to N-HtM.
  • Towards Stimulus Policy Generation
    <quote>

    • First we showed that the drop in the consumption response to a fiscal stimulus payment as the size of the payment increases, is much steeper in the model that allows for W-HtM behavior than in the models that do not.
    • Second, we showed that the model that allows for W-HtM behavior implies that to maximize the aggregate consumption response to fiscal stimulus payments, the payments should feature more moderate phasing out with household income.</quote>

Framing

  • Life-Cycle Permanent Income Hypothesis (LC-PIH)
    • standard model
    • buffer-stock model
  • the elasticity of inter-temporal substitution; is not zero
    • implies a break-down of the forward looking Euler equation holding with equality.
  • hand-to-mouth consumers (HtM)
  • marginal propensity to consume (MPC)
  • Other Frameworks (that miss)
    • toy studies; frameworks either feature only one asset or feature two assets with different risk profiles, but with the same degree of liquidity
    • Bewley models with uninsurable idiosyncratic risk and credit constraints
    • spender-saver models; patient-vs-impatient consumers; complete markets
  • Other Studies (of the same)
    • <quote>Lusardi, Schneider and Tufano (2012), who document that nearly one half of U.S. households would probably be unable to come up with $2,000 in 30 days.</quote> page 32.
  • Two-Asset Models
    • Classes
      • poor hand-to-mouth (P-HtM)
      • wealthy hand-to-mouth (W-HtM)
      • not hand-to-mouth (N-HtM)
    • Claim:
      • W-HtM are a distinct third class because…
      • W-HtM spend like P-HtM
      • W-HtM have demographics like N-HtM in {all,some,enough} other respects.
  • The Model
    • The two-asset portfolio
    • A two-period timeframe
    • Certain parameter configurations, a portfolio composition with positive amounts of illiquid wealth and zero liquid wealth is optimal.
    • W-HtM in that configuration are better off bearing the welfare loss from
      income fluctuations rather than smoothing consumption.
    • Assumption: there exists a long-term high-return high-illiquid (high transaction cost) asset class.
  • Robustness Questions
    (in majoritarian-significant enough numbers for us to care)

    • Do any of these people even exist?
    • Do they exist over significant periods of time?
      Is it a transient or persistent state (a lifestyle).
  • Data Sources
    • US → Survey of Consumer Finances (SCF), since 1983, stable since 1989.
      The Board of Governors of the Federal Reserve System in co-operation with the Statistics of Income Division of the Internal Revenue Service (IRS).
    • Canada → SFS
    • Australia → HILDA
    • UK → WAS
    • Euro → HFCS
  • Measurement
    • liquid wealth → short-term cash – short-term debt
      • the imputation procedure
      • Survey of Consumer Payment Choice (SCPC)
      • ratio = Average Cash Holdings via SPC 2010 / Median Value of Cash Accounts SCF 2010
      • Huh? <quote>Average cash holdings, excluding large-value holdings in 2010 was $138. Median checking, saving, money market and call accounts in the 2010 SCF is $2500, making the ratio about 5.5%. </quote> footnote, page 24.  How is this relevant to W-HtM?
      • The “median” of all other liquid wealth is infinitesimal-to-zero.
    • illiquid wealth → a house.

Findings Claimed

  1. HtM → 25-40% of households [US]
    • P-HtM → 33% of that → 8-12% of all housefholds
    • W-HtM → 66% of that → 13-25% of all households
  2. HtM → is age correlated
    • P-HtM are under 40 “young”
    • W-HtM peak at 40 “hump shaped at 40″
      • Figure 5, page 34
      • age span 20-80
  3. W-HtM → Something about “sizable amounts of wealth” → $50,000 in illiquid assets at age 40
    • Claim unclear.
    • This is an uncontextualized, unnormalized [US] number
    • Contextualized page 34; high income is $70,000/year for N-HtM.
    • Contrast with
      • electronic gear (nationally normed)
      • travel [plane] flights,
      • labor service rates [e.g. doctor, dentist, accountant, auto mechanic, gardner, etc.]
      • car prices (nationally normed),
      • house prices (locally normed).
  4. W-HtM are like N-HtM in other attributes
  5. Transience → yes
    • See Table 4, page 37; the numbers don’t add to 100%
    • N-HtM is an absorbing state;
    • P-HtM is semi-persistent
      • 2-year timeframes
      • P-HtM → P-HtM is 52%
      • P-HtM → N-HtM is 37%
      • P-HtM → W-HtM is 11% (apparently; 100%-52%-37%)
    • W-HtM is transient
      • lasting 2-5 years; median 29 months
      • W-HtM → W-HtM is 17%
      • W-HtM → N-HtM is 83% (apparently; 100%-17%)
  6. 50% of HtM consumers are “missed” by “Net worth” estimation
    How is this allocated?

    • all within W-HtM
    • evenly between P-HtM and W-HtM

Actualities

Via: backfill