1 Consumption and Housing2Contents 1. Core Definitions............................................................................3 2. A Simple Model of Consumption.................................................631. Core Definitions • Total wealth: • Human wealth + Non-human wealth • Human wealth: Estimated present value of after-tax labor income likely to be over the span of his working life. • Non-human wealth: • Financial wealth + housing wealth. • Financial wealth: • Stocks, bonds, checking accounts, saving accounts (which is also the estimated present value of after-tax financial income, e.g., dividends from stocks (=S), interest rate from bonds (=B)) • Housing wealth: • Value of owned-house minus mortgage still due. • How do people decide how much to consume and how much to save? • In our first model, we ignored the effect of wealth on this decision. • Now, we would like to make more accurate description and characterization recognizing effects other than current disposal income. • Notice that consumption accounts for the cheer size of GDP, so it is important to be more accurate.4• A rational consumer would: • Try to smooth her future consumption based (positively) on her expected total wealth. • Increase large purchases (i.e., housing) when finance is cheaper. • Account for the “free” retirement income from employer or government. • Increase saving (decrease consumption) when uncertainty/worry about health and life expectancy. • Increase saving (decrease consumption) to satisfy his desires for leaving bequests. • There are two economic theories (hypotheses) that recognize and endogenize some of the above effects: • Life Cycle Hypothesis, by Franco Modigliani from MIT, emphasizes that consumers’ natural planning horizon is their entire life. • Permanent Income Hypothesis, by Milton Friedman from Chicago, emphasizes that consumers look beyond current income.5• Life Cycle and Permanent Income Hypotheses both: • Recognize current consumption choices reflect thinking about lifetime income (their total wealth rather than income only) and spending. • Consumers’ spending should be smoothed from year to year (rather than vary widely from year to year). • Predict short-run MPC < APC. • Expect dissaving in retirement years. • Note: • Is the age distribution of population (Demographics) important? • How expectation for future higher output affect today’s consumption? (Hint: how will this affect future wages and dividends?) • How do liquidity constrains affect the consumption spending decision under the Life Cycle and Permanent Income Hypotheses? How does their relaxation affect the consumption spending decision? • How would a tax change affect the consumption? (hint: does that depend whether it is finance through change in G or government debt?). Barro theory (Ricardian equivalence or Ricardo-Barro proposition) intriguing but does not fair well empirically (Consider how lagged responses create a multiplier that changes over time).62. A Simple Model of Consumption • A realistic model would account, inter alia, for: • Lifecycle restrictions • Liquidity restriction • Uncertainty and risk aversion • Bounded rationality • Therefore, based on the aforesaid, a simple model would be of the form: Ct = C(Total wealtht, Cost of Consumptiont, expected remaining life) • Or: Ct = C(YDt, Total future wealtht, Pt, rt, expected remaining life) • When real interest rate is high, then it means that- - Today’s consumption leads to a greater forfeit of future consumption - Higher motivation for saving (lower for dissaving); - Lower present value of future wealth.7• Assuming P is given (zero inflation), therefore, r=i and P affects the intercept. • Therefore, a good specification would be of the following: Ct = c0 + c1 YDt + c3 TWt+1 • Or, another helpful characterization would be: Ct = c0 + c1 YDt + c4 YDt-1 + c5 dtTW - Assuming a given interest rate and prices; - Fixed effects (current and future incomes and interest rate) are showing up in the intercept; - The changes from last year are what motivate the changes in consumption (the PIH). • In the long-run (steady-state), YDt=YDt-1 and dtTW=κYDt, therefore: C = c0 + (c1 + c4 + c5 κκκκ) YD8• In the long-run (steady-state), YDt=YDt-1 and ∆tTW=κYDt, therefore: C = c0 + (c1 + c4 + c5 κκκκ) YD • Therefore, we can define two different Propensities to Consume: • Average Propensity to Consume (APC): Level of Consumer Spending/ Level of Disposable income- Ct/YDt (≡ c0/YD +MPC) • Marginal Propensity to Consume (MPC): for any specific time interval = Absolute change in spending / Absolute change in disposal income- dtC/dtYD (≡ c1 for the short run, and ≡ c1 + c4 + c5 κ for the long run). • Elasticity of Consumption (ηC,Y): Percentage change in spending / Percentage change in disposal income (≡ RC/RY=[dC/∆Y] / [C/Y] = MPC / APC ). • So if “autonomous consumption≡ c0” is small enough, then the long-run MPC=APC, and therefore, the LR elasticity is approximately 1. • In the “Long-Run”, both APC and MPC appear to be close to 95% for the US in the postwar period.9Average Propensities to Consume80%82%84%86%88%90%92%94%96%98%1967196819691970197119721973197419751976197719781979198019811982198319841985198619871988198919901991199219931994199519961997199819992000% of Disposable Income0%10%20%30%40%50%60%APC-Total APC-Durables APC-Nondurables APC-ServicesNote: 1) Different trends and 2) Differential cyclicaltitesTotal --- graphed on left scale; Components--graphed on right scale10 Obvious Wealth Efffects:APC charted versus Wealth/Income Ratio80%82%84%86%88%90%92%94%96%98%19671969197119731975197719791981198319851987198919911993199519971999APC3.54.04.55.05.56.06.57.07.5Wealth/Income MultipleAPC-Total Assets / Income11 MPCs : Annual Changes in C and YD$(50)$-$50$100$150$200$250$300$35019681970197219741976197819801982198419861988199019921994199619982000Real disposableTotal CServicesNondurablesDurables• Durables are less volatile than income • Durables are less volatile than
View Full Document
Unlocking...