Transcript Generational Aspects of Fiscal Policy under Changing Demographic

Generational aspects of fiscal policy
under changing demographic
forecasts
Jukka Lassila
(ETLA)
MoPAct workshop
Helsinki, June 2016
Motivation and aim
• In an economy with ageing population, we
analyze long-run fiscal strategies, with a novel
method.
• (fiscal strategy – a set of rules and practices
describing how policy parameters change)
• What does this method tell us about
intergenerational fairness?
• The ultimate aim is to find strategies that could
be described as both fiscally sustainable and
intergenerationally fair.
Intergenerational equity
1. between contemporary generations
2. in public transfers between generations
3. in private transfers between generations
4. between contemporary and future
generations, as yet unborn
Piachaud, Macnicol and Lewis (2009)
Intergenerational equity in public
transfers and services
• PAYG pensions are a good reference because
– young people pay, old receive, and cohort sizes
vary with aging population
– pension policy changes always(?) hit some cohorts
more than others.
• Many similarities in other public sector areas,
but also differences
– issues unrelated to age (administration, defence)
– age-neutral instruments (VAT, municipal taxes)
– stabilization issues
Previous studies
• Jensen, Nødgaard and Pedersen (2002): Tax smoothing
v. debt smoothing
• Miles and Iben (2000): Winners and losers from
pension reform, bequests
• Auerbach and Lee (2009): generational uncertainty and
risk sharing in pension systems, in context of economic
and demographic uncertainty, with the goal of finding
properties that do not result from some particular
demographic circumstances.
• Elder (1999): Individuals do not know the duration of a
tax cut with certainty. Revised expectations and
optimization. Focus on capital formation (closed
economy).
Auerbach and Hassett (2001)
• ”…how and when to deal with long-term fiscal
imbalances that are at once very significant and
very uncertain,…”
• ”…stylized models in search of more general
conclusions regarding the nature of optimal
policy responses.” (OLG with 2-period lives)
• Here: Finnish public institutions, OLG with 16period lives
• Starting point: Lassila – Valkonen – Alho, Int. J. of
Forecasting 2014
Alho: Forecasting demographic forecasts
Int. J. of Forecasting 2014
• “We assume that an approximation to the
predictive distribution of the future population is
available in terms of simulated population counts.
• The required conditional expectations are then
obtained by averaging the future evolution of a
set of sample paths that come from the
neighborhood of a target path.
• This is formally equivalent to n-nearest neighbor
kernel regression.”
The economic model
• Open-economy Auerbach – Kotlikoff type general
equilibrium model
• Overlapping generations, life-cycle optimization
• Forward-looking firms, maximizing share value
• Perfect foresight, except for demographics where
forecasts are believed in
• A revised demographic forecast every 5-year period.
Households and firms re-optimize.
• Re-optimization leads to (perceived) changes in
welfare
Demography-related features in the
model
• Earnings-related pensions (DB, with longevity
adjustment)
• and their funding (partial)
• National pensions, means-tested on earningsrelated pensions
• Health and LTC costs, depending both on age and
proximity to death
• Public education costs and some transfers
• Aggregation: How many people in what age doing
what (working or not, consuming, saving, paying
taxes, getting transfers etc.)
Welfare changes under which
strategies?
• Base strategy (”current policy”)
• Base strategy with conditional VAT changes
– Raise VAT by 2% if forecasted debt/GDP exceeds 60 %
within 20 years
• Tax smoothing policies
– Based on forecasts, periodically revised
– With varying smoothing horizons
Base strategy
• Mandatory pension contributions adapt to pension
expenditure
• Health and long-term care costs are financed partly by
municipal taxes, which adapt.
• Part of health and LTC is financed by state aid to
municipalities. State tax rates are held constant, so
changes in expenditure and tax bases go into public
debt.
• Consolidation after 55 years (smoothing taxes from
2068-72 to 2143-47)
Tax smoothing policies
• Government sets a constant VAT rate so that the state’s
debt/GDP gets back to initial level after, say, 50 years –
if the population forecast comes true. VAT rate updated
when new forecast arrives.
• Municipalities: constant income tax rate, debt/GDP
back to initial level after 50 years – if forecast is good.
Updated every period.
• Earnings-related pensions: constant contribution rate,
funds back to normal level after 50 years , updates
• Consolidation after 50 years (smoothing taxes from
2063-67 to 2143-47)
Welfare changes under base strategy,
50 % predictive limits
CV, % of consumption
40-44
from
revisions
-1.01 - +1.08
80-84
-4.17 - +3.62
Welfare changes under base strategy,
50 % predictive limits
CV, % of consumption
40-44
from
revisions
80-84
-1.01 - +1.08
-4.17 - +3.62
from
consolidation -4.82 - -2.93
-5.49 - -3.24
Welfare changes under base strategy and
tax smoothing, 50 % predictive limits
CV, % of consumption
Base
50 yrs
Base
50 yrs
from revisions
40-44
80-84
-1.01 - +1.08 -4.17 - +3.62
-1.49 - +0.97 -4.79 - +3.50
from consolidation
-4.82 - -2.93
-5.49 - -3.24
-2.16 - +0.69 -2.37 - +0.56
Welfare changes under tax smoothing
50 % predictive limits
Smoothing
horizon
20 yrs
50 yrs
20 yrs
50 yrs
CV, % of consumption
from revisions
40-44
80-84
-1.55 - +1.08 -4.85 - +3.70
-1.49 - +0.97 -4.79 - +3.50
from consolidation
-4.55 - +1.68 -4.88 - +1.52
-2.16 - +0.69 -2.37 - +0.56
Which intergenerational aspects does
our method address
1. between contemporary generations
- revision effects by age, consolidation effects by age
2. in public transfers between generations
- yes, next slide
3. in private transfers between generations
- not dealt with
4. between contemporary and future generations
- comparing revision effects to consolidation effects
Should strategies
include client
fees?
Should strategies
include client
fees?
Pre-announced
client fees?
Should strategies
include client
fees?
Pre-announced
client fees?
Prefunding?
Some observations on smoothing
• Forecast-based tax smoothing is imperfect, sometimes grossly so.
• Still it probably improves intergenerational fairness.
• With smoothing, public debt is well in control. However, the EU’s debt rule
and other fiscal rules may be violated.
• In practice only few speak about or do smoothing, let alone 50 year debt
targets.
– Canadian pension system (CPP) uses forecasts explicitly, and goes to great
length to reduce doubts concerning forecast manipulation.
– Swedish NDC system ignores forecasts completely.
– Finnish earnings-related pension system (TyEL) aims at smoothing
• Municipalities in Finland cannot take debt to any significant amount. They
could (but don’t) prefund for future health and LTC expenditure.
• With aging populations, tax smoothing strategies cause tax increases.