Quantitative Public Economics

1
Quantitative Public Economics

• Dr Hamish Low
• Lecture 4
• March 2009

2
Supervisions

• All supervisions in C5 New Court, Trinity
• Not restricted to 6 per group
• Questions from the course outline

3
Outline Evaluating Fiscal Stimulus Packages

• Examples
• Theoretical analysis
• Bush cut of 2001
• qualitative analysis
• estimating the MPC
• Darling tax cut changing relative prices

4
Examples

• Bush tax cuts of 2001
• cut in income tax rate from 15 to 10 for part
  of lower income range
• and 2008
• Obama tax cut of 2009
• Darling VAT cut of 2009
• cut in VAT from 17.5 to 15 until the end of 2009

5
Simple Theory A temporary income tax cut

• Increase in consumption depends on income
  elasticity

6
A revenue equivalent permanent tax cut

• Increase in consumption depends only on change in
  lifetime income
• The timing of the income change is unimportant
  (if no uncertainty)

7
Importance of Liquidity Constraints

• Liquidity constrained increase consumption 1
  for 1 extra income

8
Consumption smoothing
(1r) is the relative price of consumption in t
compared to period t1
(Homothetic, so consumption in t increases
proportionally with lifetime income)
Consumption growth higher if positive expectation
error
9
A change in relative prices (eg through VAT)

• Increase in consumption depends on
• change in lifetime income
• change in relative prices

Effect of change in intertemporal prices is given
by the slope of the indifference curve the
elasticity of intertemporal substitution
10
Labour Supply Effects

• Change in the income tax rate have usual income
  and substitution effects on labour supply.
• For most workers, there is only an income effect,
  operating to reduce labour supply
• Intertemporal substitution effect on labour
  supply tax rate lower in t than in t1 and so
  work more in t than in t1.
• Labour supply intertemporal substitution thought
  to be small

11
Summary

• Fiscal stimulus through cutting taxes
• how much does lifetime income change (permanent
  vs transitory tax cut Ricardian equivalence)
• how responsive is current consumption to lifetime
  income
• liquidity constraints
• if tax cut through prices, response of
  consumption depends also on elasticity of
  intertemporal substitution

12
Bush 2001 Tax Rebate

• US economy downturn after the DotCom bust
• 300 to 600 tax rebate sent out in 3rd quarter
  2001
• Arose because
• previous 15 tax band split into 10 on the lower
  part (6000 for singles), and 15 on the
  remainder
• backdated to Jan 1st 2001, so rebate necessary
• Value 38 billion, 0.4 GDP

13
Bush 2001 Tax Rebate

• Federal policy no control group
• But, randomisation in the timing of the rebate
• Two questions
• what is the marginal propensity to consume out of
  a tax cut (temporary increase in income)?
• is the announcement of the tax cut or the receipt
  of the money more important?

14
Path of Consumption Through the 2001 US Recession
15
Alternative Approaches

• Time series data
• Questionnaire
• Look for exogenous variation / randomisation in
  the rebate
• Model of consumer behaviour

16
Shapiro and Slemrod (AER, 2003)

• Qualitative questions
• The tax rebate will be 300 for single
  individuals and 600 for married couples.
  Thinking about your familys financial situation
  this year, will the tax rebate lead you mostly to
  increase spending, mostly to increase saving, or
  mostly to pay off debt?
• Other questions on (i) expectations about future
  tax cuts (ii) implications for government
  spending
• Gives idea of whether expect it to be permanent
  or transitory

17
Unconditional Responses

• No evidence that lower income households more
  likely to spend
• Translates into an Marginal Propensity to Consume
  out of the rebate of about 0.33

18
Johnson, Parker and Souleles (AER 2006)

• Wealth effect of rebate occurs at the time of the
  announcement.
• Wealth effect uncorrelated with the timing of
  receipt of the rebate
• Key is that announcement happened in period t,
  but rebate arrived later in t1
• Amount of the rebate is not random

19
Change in Expenditure
For singles, rebate equals 300, so total
effect 33 increase in food, 72 increase in
total ND
20
Change in Expenditure
21
Cumulative Effects Fraction of Rebate Spent
22
Identification

• Cross-section variation in the timing of the
  rebate
• Should not use variation in the amount of the
  rebate
• those with higher rebate have higher permanent
  income
• Other evidence shows rise in payments of credit
  card debt

23
US Rebate of 2008

• Between 300 and 600 for singles between 600
  and 1200 for married couples
• Temporary (whereas in 2001, was part of a 10 year
  plan)

24
Shapiro, Slemrod (NBER Working Paper, Feb
2009)Qualitative Questions

• Very similar to 2001 responses

25
The Effect of the Rebate on Saving
Saving Rate
Saving Rate without rebate
26
(No Transcript)
27
UK Prospects
Source IFS Green Budget
28
Temporary Cut in VAT introduced in 2008

• VAT cut from Dec 1st 2008 until Dec 31st 2009
• Want to predict consequences for consumption
• Size of income effects
• Intertemporal substitution effects
• Revenue and Distributional consequences
• 12.5 billion, 0.8 of GDP
• 440 per household

29
Effect on Prices?

• Only 55 of goods subject to VAT
• 2.5 cut in VAT ? 1.2 fall in relative price of
  consumption today compared to 2010
• Less of a fall if less than full-pass through (eg
  imperfect competition)
• Most goods subject to VAT are luxuries or
  durables
• TVs, washing machines, cars
• clothes, restaurant meals, leisure activities

30
The effect of a change in relative prices
31
Similar to an interest rate cut

• Interest rate cut changes intertemporal prices
• Income effects differ interest rate cut
  penalises savers, rewards borrowers
• Interest rate affects investment and the
  exchange rate

32
Intertemporal Substitution

• Estimate EIS parameter using Euler equation
  (structural)
• Estimates use non-durable consumption, usually
  ignore luxuries vs necessities distinction
• Attanasio and Weber (JPE 1995) EIS 0.67
• Perhaps this is a lower bound luxuries and
  durables have greater intertemporal substitution
  possibilities

33
Predictions

• EIS 1 ? all of the extra income is spend (MPC
  1)
• For the liquidity constrained, the MPC is also 1
• Consumption 1.2 higher in 2009 than otherwise
• difficulty of identifying the counter-factual
• 12.5 billion revenue spent increases consumer
  spending by 12.5 billion
• 0.8 increase in GDP

Source Crossley, Low and Wakefield, 2009
34
Distributional Consequences

• Rich benefit most from the VAT cut because they
  pay proportionally more in VAT a progressive tax

35
Conclusions

• Alternative approaches to evaluating fiscal
  stimulus packages
• After the event time series not informative,
• use questionnaires
• try to find randomisation and estimate response
• In making predictions use economic theory and
  estimates of preference parameters
• Fiscal stimulus through temporary tax cuts likely
  to be effective if changes relative prices