Transcript The Great Moderation and 'Falling Off a Cliff': neo

“The economy fell off the cliff.”
– George Soros (11/24/2008).
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
1
The Great Moderation and "Falling
Off a Cliff": neo-Kaldorian dynamics.
James G. Devine
Loyola Marymount University (LA, CA)
[email protected]
August 5, 2011
(Published as The Great Moderation and “Falling Off a Cliff”: Neo-Kaldorian Dynamics in
Journal of Economic Behavior & Organization, 78(3), May 2011: 366-373. A rough draft
is available at: http://myweb.lmu.edu/jdevine/JD-2010-neoKaldorianModel.pdf with
diagrams at http://myweb.lmu.edu/jdevine/neoKaldorian-Figures.docx. The current
version has been edited a lot without changing any conclusions.)
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
2
Outline.
I. Introduction.
II. The Short-Run Model.
A.
B.
C.
the Expenditure curve (EE).
the Actual Demand/Debt ratio curve (AA).
Short-run equilibria.
III. Medium-Run disruption of SR Equilibrium.
A.
B.
C.
The “typical” cycle.
A Fall off the cliff.
The Great Moderation.
IV. The aftermath: Recovery or Stagnation?
V. Conclusion: Policy’s role.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
3
(I) Purpose.
to help to understand:
Why the U.S. economy “fell off a cliff” –
or threatened to do so – during the
years 2008-2009.
the world economy, the housing market, and
most of finance will be ignored here.
it’s possible that 2009’s stimulus package saved
the economy from a Fall – but who knows?
o We may not have been suffering from a Fall.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
4
(I) Theoretical Background.
1.
Kaldor’s Keynesian (pre-catastrophe theory) model of the
business cycle (1940):



2.
Non-linearity implies three equilibria (two of which are stable).
Stable equilibria represent two general states of the macroeconomy: high employment and stagnation.
A “Fall” is a downward leap between these.
Dynamic theories of Minsky (1982) and Kalecki (1933),
helping to cause this Fall endogenously.


This process may have occurred due to the often-heralded “Great
Moderation” (1984-2006).
In this period, the effects of financial crises and normal businesscycle recessions may have been short-circuited, so that they could
not purge the economy of Minsky/Kalecki imbalances.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
5
(II) The three “runs.”
• long run (LR): labor-constrained potential output
(Z) and the Minsky/Kalecki threshold (V) are
determined.
 Assumed constant in this paper.
• medium run (MR): the trend demand/debt ratio (t)
and the Spending shift factor (St) are determined –
but are held constant in the short run.
• short run (SR): Bt (private-sector debt) and Kt
(industrial capacity) held constant in the short run.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
6
(II) The Subject Matter.
Figure 1
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
7
(II)
Three SR Equations.
1. The EE (expenditure) curve relating demand for
GDP (Et) to the expected demand/debt ratio (xt);
2. The AA line, determining the actual demand/debt
ratio (at) at each level of demand (Et); and
3. Expectations adjustment, so that the expected and
actual demand/debt ratios are equal in short-term
equilibrium (x = a).
 The adjustment equation is left implicit here.
 It’s treated as merely a matter of an automatic
movement to short-run equilibrium.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
8
(II.A)
Aggregate Expenditure (EE).
Figure 2
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
9
(II.A) Short Run: movement along
EE curve.
Et = EE(xt, St); EE1 ≥ 0; EE2 ≥ 0
(1)
• Shift Factor (St) is constant in the short run.
• The sigmoid shape of the EE curve:
1. Between the two vertical segments, spending
rises with the expected demand/debt ratio.
2. But investment and total spending do not respond
to xt at low demand (due to extreme pessimism,
indebtedness, and unused industrial capacity)
3. Nor at high demand (due to supply-side
bottlenecks).
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
10
(II.B) Actual Demand/Debt Line (AA).
figure 3
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
11
(II.B) The Actual Demand/Debt Line (AA)
at ≡ (Et/Z)(Z/Kt)/(Bt/Kt) ≡ et · t /λt
(2)
• The actual expenditure/debt ratio depends
on three ratios:
Expenditure/potential = employment rate (et);
The Kalecki factor: Potential/industrial capacity
(Z/Kt = t); and
The Minsky factor or the degree of leverage:
Private debt/industrial capacity (Bt/Kt = λt).
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
12
(II.B) Simplifying…
Let t /λt ≡ αt so that:
at ≡ et·αt
(2A)
• The actual demand/debt ratio (at) reflects
 the utilization of potential (et); and
 αt ≡ the trend value of at (held constant in the SR)
which reflects:
 the Kalecki factor (t); and
 The Minsky factor (λt).
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
13
(II.B) Short run: movement along AA curve.
• Since Bt and Kt are held constant in the
short run,
• λt (the leverage ratio), ρt (potential
output/capital ratio) and αt, the trend
output/debt ratio are constant.
• The actual demand/debt ratio(at) varies
only with the utilization of labor (et): a
linear relationship.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
14
(II.C) Short-run equilibria.
at = xt,
so that Et(at, St) = Et(xt, St)
(3)
(3A)
• The process of adjustment of expectations (xt)
to actual values (at) indicates that
equilibria L and H are stable, while
M is unstable.
• In figure 4, the small arrows show the direction
of disequilibrium adjustment.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
15
(II.C) Short-Run Equilibria.
figure 4
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
16
(III) Medium Run: Shifting EE curve.
• St changes and EE shifts due to
fiscal and/or monetary policy,
changes in expected inflation, and/or
changes in long-term profit expectations.
• Stimulus (the shift to EE’) means that a
lower x than before is associated with the
same amount of expenditure.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
17
(III) Medium Run: Shifting EE curve.
• The limits to Stimulus.
Near Z, the curve cannot shift to the right (only
downward) due to labor-supply constraints.
and demand-side stimulus can only be
temporary (since only inflation results in the
end).
• to describe the “Great Moderation,” St is held
constant with a high Et – indicating effects of
the trend underlying EE fluctuations.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
18
(III) Minsky/Kalecki Dynamics.
• Persistent high Et above Minsky/Kalecki threshold V implies that
either or both:
 leverage (λt) rises (Minsky).
o Extended prosperity encourages more and more borrowing as prosperity
is expected to continue and memories of the Great Depression and past
financial crises fade.
o This assumes that the financial system is poorly regulated.
 the potential-industrial capacity ratio (t) falls (Kalecki).
o High demand encourages fixed investment while the fixity of Z means that
effective “capital productivity” (Z/Kt) falls as not all of the industrial
capacity can be used to produce.
o Persistent high demand may cause “disproportionalities” as the wrong kind
of investment is done, given the structure of demand.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
19
(III) Minsky/Kalecki Dynamics.
αt = –M(Et – V); with constant M > 0
(4)
• Persistent high Et above Minsky/Kalecki
threshold V implies that
λt rises and/or t falls.
And αt falls, flattening AA, as in figure 3.
• Going the other way: persistent Et < V
rotates AA counterclockwise.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
20
(III) The Economist’s Holy Grail.
• In theory, medium-run equilibrium
exists where Et = V (with constant α).
• But can this holy grail be both attained
and maintained?
• The answer depends on the
relationship between V and the AA/EE
tangency point T (introduced below).
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
21
(III) Endogenous Disequilibration.
• Equation (4) and Et > V imply falling αt in
the MR, which
leads to endogenous disruption of any shortrun equilibrium attained.
• This in turn implies either
A. a “mild” recession or
B. a Fall off a Cliff.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
22
(III.A) A “Mild” Recession.
figure 5
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
23
(III.A) Mild Recession & Cycle.
• Holding EE constant, falling αt leads to clockwise
rotation of the AA line to AA2.
• Because the AA/EE tangency point HM at Et = T is
below the threshold V,
 the recession (declining Et) causes endogenous reversal
of decline in αt as soon as Et < V.
 Spending recovers as AA rotates counterclockwise.
• A “typical” cycle involves repeated clockwise and
counterclockwise rotation of the AA line
 … along with a lot of other considerations such as the
inventory cycle.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
24
(III.A) MR equilibrium maintained?
• Attainment of MR equilibrium can occur (at H2).
• This is a stable SR equilibrium with constant αt.
• However, this MR equilibrium requires maintenance of
relatively high unemployment of labor to prevent the
Minsky/Kalecki trend.
 This is a “reserve army of labor” theory (à la Marx).
• Standard business cycle theory suggests reasons why
the economy might oscillate around MR equilibrium.
 Nonetheless, this equilibrium is stable.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
25
(III.B) Falling Off A Cliff.
figure 6
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
26
(III.B) The Fall.
• Holding EE constant, falling αt again implies
clockwise rotation of AA to AA2.
• In this case, Et at the tangency point T is above the
threshold V.
 The same result occurs with equality of these two points.
• Thus, the recession causes points H and M to
converge to the tangency point HM, which is
unstable downward.
• Because V is low, αt continues to decline.
• So even if equilibrium at HM is maintained, the SR
equilibrium point disappears entirely.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
27
(III.B) The Fall and MR equilibrium.
• The medium-run equilibrium at Et = V
cannot be attained because it does not
correspond to a stable SR equilibrium.
• The model instead implies a Fall to point L
(stagnation).
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
28
(III.B) Does a Fall occur?
• We cannot say a priori what the relationship
between points T and V is in the real world – so we
can’t say which of the two cases occurs.
• But T is likely to be relatively high due to an
extended period of relative prosperity (such as the
“Great Moderation”) which allows imbalances to
accumulate, lowering αt for long periods.
• With T associated with a higher level of Et, a Fall is
more likely.
• This kind of trend is seen in figure 7, even if the
“Moderation” was anemic from labor’s perspective.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
29
(III.C) The Great Moderation and Falling αt.
figure 7
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
30
(III.C)
Illustrative regression: the trend.
ln(at) = 0.3704 – 0.0027·(time index)
Adjusted R2
0.5848
Standard Error
0.0623
95
(Great Moderation only)
Coefficients
t-stat
Constant
0.3704
27.0460
Time coefficient
–0.0027
–11.5499
Observations
Data Source: Federal Reserve Flow of Funds accounts.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
31
(III.C) A Major Caveat.
• Even though the trend in t is statistically
significant, that does not mean that we can
conclude that the fall was large enough to
explain the 2008-9 collapse of the U.S.
economy.
• To say that would require that we know
much more about the shape of EE and the
structure of the economy.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
32
(V) Recovery or Stagnation?
figure 8
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
33
(IV.A) the aftermath & Recovery.
• In figure 8, because Et at point L3 < T, there
is an automatic tendency toward recovery
due to deleveraging (λt) and purging of
unused capacity (ρt ).
• So αt rises, rotating AA counterclockwise.
• Equilibrium points L and M converge to ML,
which is unstable upward: the economy
leaps up the cliff.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
34
(IV.B) the aftermath & Stagnation
• Et at point L might exceed T, so the Minsky/Kalecki
trend continues, making matters worse.
• More importantly, recovery can be counteracted by
the MR results of extreme unused capacity and
indebtedness, which encourages waves of deflation,
default, and rapid-onset despair.
 These shift EE up and left to EE’: higher x is required to
induce the same amount of spending as before.
 Continued or deepening stagnation results.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
35
(V) Policy’s Role.
• Policymakers can be “villains” by maintaining
high demand – encouraging the accumulation
of imbalances as in a Great Moderation.
• But in a stagnation period, they can become
“heroes” by stimulating demand.
Fiscal policy (if politically possible) and monetary
policy (if it works) can “prime the pump,” spurring
recovery.
This shifts EE downward, moving the tangency
point to the left, making recovery more likely.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
36
(V) To Create a New Prosperity.
• What encourages the creation of new
prosperity?
1. Efforts to lower the leverage ratio (λt), via
mass bankruptcy and the like.
2. Efforts to raise “capital productivity” (ρt) by
scrapping excess and/or inappropriate
industrial capacity.
• Both of these artificially raise αt rather
than waiting for the “automatic” process.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
37
Preventing falling αt
• Increased leverage can be prevented using
improved financial regulation.
• Decreased capital productivity cannot be
prevented without raising Z or avoiding
disproportionalities.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
38
(V) To Preserve the New Prosperity.
• What can encourage the persistence of new
prosperity, preventing the negative effects of
a new “Great Moderation”?
1. Raise Z by increasing the supply of labor.
2. Raise Z by increasing labor productivity.
• If successful, both of these allow
persistently high demand by increasing
supply in step.
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
39
Finis
8/5/2011 rev.
J. Devine/Neo-Kaldorian Dynamics
40