Party Without A Hangover by Douglas Laxton

1
Party Without A Hangoverby Douglas Laxton

• Comments
• by
• Jeffrey Sheen
• University of Sydney
• CAMA Conference on Fiscal Policy Frameworks
  Monetary Policy Implications, and
  Intergenerational Financial Funds
• Thursday 23 and Friday 24 August 2007

2
Main questions

• 1. What were the sources of the huge US current
  account deficits over the last decade?
• 3 primary sources
• A decrease in public and private sector saving in
  the US (which would raise global real interest
  rates)
• An increase in saving in the rest of the world
  (which would lower global real interest rates)
• An increase in relative investment in the US due
  to its productivity growth advantage.
• This paper analyzes the first two only.
• 2. What differences arise with changes in
  unproductive and productive government
  expenditure?
• Using a heavy non-Ricardian model, it gets
  (mostly) plausible results

3
Key features

• A SERIOUSLY LARGE COMPLEX MODEL
• 2 countries (US and ROW)
• New Keynesian /New Open Economy Macro
• OLG with
• optimizing households Forward-looking plus
  liquidity-constrained
• birth, death retirement
• disconnected cohorts, finite horizons declining
  productivity
• Endogenous labour supply
• no unemployment
• No intergenerational bequests or gifts
• Perfect life insurance with annuities
• External habit formation
• Perfect foresight
• Incomplete international asset markets
• only bonds are traded (not equities)
• and thats just the consumers!!!

4
Key Features cont.

• Production and Sales
• (with ubiquitous Dixit-Stiglitz varieties
  price-quantity adjustment costs, yielding
  multiple Phillips curves)
• Manufacturing sector
• traded and non-traded
• CES in capital labour
• Unions (really, employment agencies)
• profit by setting wages to buy labour from
  households sell to manufacturers
• Import agents
• buy traded goods from other country
• Distributors (maybe, wholesalers)
• buy domestic goods and also foreign goods from
  import agents, combine them with
  government-provided infrastructure capital to
  produce retail goods
• Retailers
• combine the above and sell to households

5
Key Features cont.

• Government
• Fiscal Policy
• Distortionary labour and consumption taxes
• Rule for transfers to redistribute income between
  households
• Wasteful government consumption plus productive
  investment in public infrastructure
• Taxes adjust to achieve a target for fiscal
  surplus/GDP.
• Monetary Policy
• Nominal interest rate rule responding to an
  inflation target and output gap growth
• Non-stationary steady state real interest rate
• Money plays no significant role (cashless
  limit)

6
Sources of the Non-Ricardian Results

• In general government deficits/taxes/debt matter
  if there are
• Non-altruistic generations, or else sub-optimal
  bequests
• Non-distorting taxes levied on different people
  at different times.
• Birth of disconnected cohorts is necessary with
  infinite or finite horizons (Weil, 1989)
• Probability of death leads to higher discount
  rates for future tax liabilities and for future
  utility (Blanchard, 1985), amplifying the results
• Lifecycle productivity changes can amplify
  results.
• Incomplete asset markets giving a proportion of
  liquidity constrained households
• Distorting labour and consumption taxes
• Endogenous full employment consumption varies
  with the tax pattern
• All of these non-Ricardian features are in the
  paper, which compares finite and infinite horizon
  models.

7
Key Results of the Paper

• Infinite horizon model has plausible though
  modest effects within a business cycle
• Finite horizon model has persistent and large
  non-Ricardian effects
• Permanent increase in government debt seems to
  lead in the long run to a
• continuous decrease in US GDP (due to higher
  income taxes) and RoW GDP
• fall in US investment
• worsening in the current account and real
  exchange rate
• continuously rising real interest rate (beyond 50
  years!)

8
Persistence Puzzle in the FH Model Higher debt
seems to lead to ever- rising real interest rate

• Shown not to be true in an infinite horizon OLG
  model (REP)i.e. the modified Weil 1989 model.
• But it was also not true for FH in Blanchard
  1985. So what in this modified Blanchard model
  is generating the persistence result?
• Could it be that the finite horizon is
  interacting with tax distortions on labour supply
  via the fiscal target?
• What happens if that channel is switched off
  (?olg1)?
• What if there were no distorting taxes?
• Could it be because money plays no role um0?
• But, why did the real interest rate converge in
  the infinite horizon model?
• Switch to REP included a risk premium in UIP
• What would happen with a risk premium in the FH
  model
• Maybe 50 years isnt enough time for convergence!
  Too many costs of adjustment!
• A coding error?

9
Some suggestions!

• Unemployment
• Labour market clears for any amount of hours
  demanded and supplied
• To get unemployment fluctuations, could add a
  cost so that individuals adjust at the extensive
  margin (work or not work), rather than at the
  intensive one (hours).
• See Gary Hansen JME 1985 eg cost of travelling
  to work. For RBC models, this delivered
  realistically large fluctuations in hours worked
  and small productivity fluctuations.
• Calvo pricing, rather than costs of adjustment
• Which would calibrate better?
• Ageing population is one main low frequency
  explanation for high current savings in RoW.
• Rather than increasing b, model this with a
  differential birth rate.
• Differential in productivity growth is one main
  high frequency explanation for CA deficit.
• Introduce relative investment productivity as a
  credible explanation of the US CA deficit.