What Would You Do With $500? Spending Responses to Gains, Losses, - PowerPoint PPT Presentation

What Would You Do With $500? Spending Responses to Gains, Losses, News and Loans Andreas Fuster Greg Kaplan Basit Zafar Reserve Bank of New Zealand December 2017

Another Paper Estimating MPCs ? • Approaches to estimating MPCs: 1. Revealed reference: natural experiments, transitory income 2. Reported preferences: hypothetical surveys • Lessons from revealed preference literature: • Average quarterly MPC: 15%-30% • Heterogenous, bi-modal, correlated with liquid wealth • Limited power for developing and distinguishing theories • Natural experiments: mostly one-time unanticipated windfalls • Correlating with observables: limited to liquid wealth 1 Fuster, Kaplan and Zafar (2017)

This Paper • Use hypothetical survey responses to test and refine existing models of consumption behavior • Within-person variation from alternative treatments: • GAIN : size effect • LOSS : sign asymmetry • NEWS about gains and losses • LOAN • Better hypothetical survey questions than existing literature • Compare with theoretical predictions from simple models • Implement treatments in calibrated life-cycle models 2 Fuster, Kaplan and Zafar (2017)

Main Findings • Responses to GAIN treatment largely consistent with existing literature • Insights from additional treatmets: 1. (i) Sign asymmetry: LOSS > GAIN ; (ii) No response to NEWS-GAIN ⇒ importance of low levels of liquid wealth 2. Response to NEWS-LOSS ⇒ not just myopia 3. No response to LOAN ⇒ not short-term credit constraints 4. Size effects ⇒ non-convexities, durables, salience may play a role 3 Fuster, Kaplan and Zafar (2017)

Outline 1. Data 2. Empirical Results Baseline: GAIN MPC Treatment: NEWS-GAIN Treatment: LOSS Treatment: NEWS-LOSS Treatment: LOAN Summary 3. Theoretical Implications Benchmark Models Precautionary Savings Models Quantitative Lifecycle Models 4. Conclusion 3 Fuster, Kaplan and Zafar (2017)

NY Fed Survey of Consumer Expectations • Online survey to elicit expectations about economic variables • Rotating panel ∼ 1 , 300 household heads • Respondents participate for up to 12 months: monthly modules • Response rate: • First-time invitees: ∼ 55% • Repeat respondents: ∼ 85% • Four additional modules asked to repeat panelists: Mar 2016, May 2016, Jan 2017, Mar 2017 • Panel nature: some respondents answered multiple modules • Total of 9,086 responses from 2,586 panelists 4 Fuster, Kaplan and Zafar (2017)

Sample Characteristics Demographics Sample ACS 2015 White/Non-Hispanic 0.75 0.69 Age 50.43 51.06 Education BA+ 0.56 0.31 Married 0.64 0.50 Homeowner 0.73 0.59 Midwest 0.25 0.21 Northeast 0.20 0.18 South 0.33 0.38 West 0.22 0.24 Financial Characteristics Sample SCF 2013 Income < = 50k 0.36 0.37 Income 50k-100k 0.36 0.30 Income 100k+ 0.28 0.31 Financial Assets > 20k 0.50 0.35 Non-housing Debt > 20k 0.35 0.23 Net Worth > 200k 0.42 0.34 5 Fuster, Kaplan and Zafar (2017)

Baseline Survey Instrument GAIN: Quarterly MPC out of one-time unexpected receipt of $ Y • Asks respondents how spending behavior would change over subsequent three months in response to unexpected increase in resources. • Three amounts: $500 , $2500 , $5000 • Elicit spending response in two stages: 1. Ask whether spending / debt payments / savings would change in response to windfall 2. Ask about magnitude of change 6 Fuster, Kaplan and Zafar (2017)

Survey Screen Shot 1 7 Fuster, Kaplan and Zafar (2017)

Survey Screen Shot 2 8 Fuster, Kaplan and Zafar (2017)

Survey Screen Shot 3 9 Fuster, Kaplan and Zafar (2017)

Reported Expenditures in the Literature • Shapiro-Slemrod, Parker-Souleles Thinking about your (family’s) financial situation this year, will XX lead you mostly to increase spending, mostly to increase saving, or mostly to pay off debt? • Japelli-Pistaferri (Italy) Imagine you unexpectedly receive a reimbursement equal to the amount your household earns in a month. How much of it would you save and how much would you spend? Please give the percentage you would save and the percentage you would spend. • Graziani-van der Klaauw-Zafar What are you doing or planning to do with the extra income? Asked to report the share planning to spend / save / pay down debt. 10 Fuster, Kaplan and Zafar (2017)

Reported Expenditures in the Literature • Christelis-Georgarakos-Jappelli-Pistaferri-van Rooij (Netherlands) Imagine you unexpectedly receive a one-time bonus from the government equal to the amount of net income your household earns in (one-month / three months). In the next 12 months, how would you use this unexpected income transfer? Distribute 100 points over these four possible uses: 1. Save for future expenses [0 …100] 2. Repay debt [0 …100] 3. Purchase within 12 months durable goods (cars, home improvement, furniture, jewelry, other durable good) that you otherwise would not have purchased or that you would have purchased later [0 …100] 4. Purchase within 12 months non-durable goods and services that do not last in time (food, clothes, travel, vacation, etc.) [0 …100] • Boon-Le Roux-Reinold-Surico (UK) If your household received an unexpected windfall of [amount] tomorrow, what do you think you would do with this extra money? • I would spend all [100%] or more this year • I would spend between [75%] and [100%] more this year • I would spend between [50%] and [75%] more this year • I would spend between [25%] and [50%] more this year • I would spend between [1] and [25%] more this year • I would not change spending and save the windfall • I would not change spending and use the windfall to pay off some of my debt 11 Fuster, Kaplan and Zafar (2017)

Advantages of Our Questions • Explicitly state size of windfall, allowing it to be varied • Two-stage approach: • Allows more precise estimates of zero MPCs • Does not prime respondents toward a non-zero response • Explicit about time horizon over which spending response is measured • Does not impose that MPC is between 0 and 1 • Richer information from alternative treatments 12 Fuster, Kaplan and Zafar (2017)

Treatments • GAIN: Quarterly MPC out of one-time unexpected gain of $ Y • LOSS: Quarterly MPC out of one-time unexpected loss of $ Y • NEWS-GAIN: Quarterly MPC out of unexpected news about a gain of $ Y to be received Z quarters from from now • NEWS-LOSS: Quarterly MPC out of unexpected news about a loss of $ Y to be received Z quarters from from now • LOAN: Quarterly MPC out of unexpected interest-free loan of $ Y to be repaid 1 year from from now • Respondents exposed to two treatments in each module • Order of treatments within a module was randomized 13 Fuster, Kaplan and Zafar (2017)

Treatments Mar-16 May-16 Jan-17 Mar-17 GAIN $500 [1,086] [594] $2,500 [543] $5,000 [361] [1,087] [594] NEWS-GAIN $500 in 3 months [362] $5,000 in 3 months [594] LOSS $500 Loss [362] [1,180] NEWS-LOSS $500 in 3 months [594] [590] $500 in 2 years [590] LOAN $5,000 Loan [544] 14 Fuster, Kaplan and Zafar (2017)

MPC Summary Share of Respondents Mean with MPC MPC | MPC > 0 Count MPC Mean Median < 0 = 0 > 0 GAIN $500 1638 0.08 0.06 0.75 0.19 0.54 0.50 $2500 540 0.11 0.06 0.66 0.27 0.43 0.40 $5000 1629 0.14 0.06 0.55 0.39 0.36 0.30 LOSS $500 1536 0.30 0.04 0.47 0.49 0.62 0.60 NEWS-GAIN $500, 3 mnth 362 -0.00 0.08 0.86 0.06 0.41 0.50 $5000, 3 mnth 594 0.04 0.05 0.81 0.14 0.31 0.30 NEWS-LOSS $500, 3 mnth 975 0.31 0.02 0.46 0.51 0.61 0.55 $500, 2 yr 589 0.14 0.03 0.68 0.29 0.52 0.40 LOAN $5000 541 0.01 0.12 0.79 0.08 0.30 0.20 15 Fuster, Kaplan and Zafar (2017)

Outline 1. Data 2. Empirical Results Baseline: GAIN MPC Treatment: NEWS-GAIN Treatment: LOSS Treatment: NEWS-LOSS Treatment: LOAN Summary 3. Theoretical Implications Benchmark Models Precautionary Savings Models Quantitative Lifecycle Models 4. Conclusion 15 Fuster, Kaplan and Zafar (2017)

Outline 1. Data 2. Empirical Results Baseline: GAIN MPC Treatment: NEWS-GAIN Treatment: LOSS Treatment: NEWS-LOSS Treatment: LOAN Summary 3. Theoretical Implications Benchmark Models Precautionary Savings Models Quantitative Lifecycle Models 4. Conclusion 15 Fuster, Kaplan and Zafar (2017)

Response to Gains • Most people do not respond: 75% • Conditional on responding, substantial MPC: Mean 54% , Median 50% • Majority of response in first month after receipt Summary 16 Fuster, Kaplan and Zafar (2017)

Size Effects • Extensive: larger gains, more respond: 19% for $500 , 39% for $5 , 000 • Intensive: larger gains, smaller responses • Extensive slightly dominates: mean MPC from 8% from 14% • Fraction on durables increases with size: 24% for $500 , 36% for $5 , 000 Summary 17 Fuster, Kaplan and Zafar (2017)

Internal Consistency $500 $5,000 • Similar distributions across survey waves • Within person consistency over time: 68% in same bin ≤ 0 , (0 , 1) , ≥ 1 Summary 18 Fuster, Kaplan and Zafar (2017)

Outline 1. Data 2. Empirical Results Baseline: GAIN MPC Treatment: NEWS-GAIN Treatment: LOSS Treatment: NEWS-LOSS Treatment: LOAN Summary 3. Theoretical Implications Benchmark Models Precautionary Savings Models Quantitative Lifecycle Models 4. Conclusion 18 Fuster, Kaplan and Zafar (2017)