2
The political structure of business cycles.
According = to [Alesina(1988)], traditional PBC theories can be organized around a four-fo= ld schema that considers “whether voters evaluate candidates retro- or prospectively, whether economic actors have adaptive or rational expectatio= ns and whether policy makers have opportunistic (office-seeking) or partisan motivations” [Franzese(2002)]. However, we should add a fifth rubric = to explain how such political cycles center in dynamic distributive tensions.<= o:p>
Traditiona=
l PBCs
explain conflicts of interest out of partisan differences or mere opportuni=
sm.
However, recent events have proven, often to the derision of much received
wisdom, that voters act out of awareness of their relative standing in the
functional distribution of income in society. Unlike Nordhaus’ canoni=
cal
model [Nordhaus(1975)] where a simple Phillips’ curve policy relation
fiddles the dimensions of inflation and unemployment, in Kalecki the policy
dimensions rest on political / economic structures around such social
distribution of income. Nordhaus’ PBCs inspired the first wave of
mainstream models (indeed Nordhaus does recognize Kalecki as an inspiration=
in
an overlooked footnote), yet subsequent work failed to spark commentary acr=
oss
distributive and productive dimensions [Olters(2004)]. Moreover, while all
these models are widely cited, researchers in the KaleckiGoodwin tradition =
have
stressed their macroeconomic implications without much discussion on their
political insights. We try to bridge this gap as a foreword to our stylized
facts.
In this
understanding of Kalecki, political leaders strive to win elections and keep
the allegiances of majority interests playing the conflict between employme=
nt
creation, deficits, wages and profits, (which take them to power and ask in
return higher wages and employment) and elite business and financial intere=
sts
(that keep them in power and profit from capital accumulation).
In countries that feature deep social
iniquities, taxes become an increasingly inconvenient alternative for
generating fiscal revenue. Therefore, governments prefer these policies to =
be
sourced by debt. Sources of government expenditure do matter, especially wh=
en
political costs to domestic stakeholders and investors factor in spending
strategies. Kalecki recognized as much when he wrote:
In the slump, eithe=
r under
the pressure of the masses, or even without it, public investment financed =
by
borrowing will be undertaken to prevent large scale unemployment. But if
attempts are made to apply this method in order to maintain the high level =
of
employment reached in the subsequent boom a strong opposition of
“business leaders” is likely to be encountered... lasting full
employment is not all to their liking. The workers would “get out of
hand” and the “captains of industry” would be anxious to
“teach them a lesson”. Moreover the price increase in the up-sw=
ing
is to the disadvantage of small and big rentiers and
makes them boom tired [Kalecki(1943)].
While
conventional wisdom emphasizes the role of such spending (via the size of t=
he
multiplier, resources provided for government action to temper fluctuations=
in
employment (either through debt spending or tax increases) are sought and
supported by different constituencies. Taxes are a politically and economic=
ally
costly source of public finance, as increased fiscal revenues dampen aggreg=
ate
demand and derivatively, election prospects for incumbents. For this reason,
governments tap financial markets to pursue expansionary economic policy.
However, such persistent government
intervention in the economy erodes a state of confidence at the core of
financial market conditions. As Kalecki recognized:
Under a laissez-faire
sy=
stem
the level of employment depends to a great extent on the so-called state of
confidence. If this deteriorates, private investment declines, which result=
s in
a fall of output and employment... This gives to the capitalists a powerful
indirect control over government policy: everything which may shake the sta=
te
of confidence must be carefully avoided because it would cause an economic
crisis [Kalecki(1943)].
This state of confidence expressed its=
elf
in many ways: it drives financial markets or nudges business plans, investm=
ent
in capital goods or business consumption. However, the binding constraint f=
or
government action is not necessarily the appetite of financial markets, but
institutional checks established to prevent popular control of economic
decisions. Indeed, dislike for government spending grows contentious if the
objects of spending are increasingly considered to rival private activities
(like public investment) or subvert the bargaining clout of firms - for
example, transfers and subsidies. But even more importantly, increased work=
er
clout (expressed by higher wages) may cause a qualitative shift in business
climate - as Kalecki explained:
Indeed, under a regime of permanent fu=
ll
employment, the ‘sack’ would cease to play its role as a
‘disciplinary measure’. The social position of the boss would be
undermined, and the self-assurance and class-consciousness of the working c=
lass
would grow... ‘Discipline in the factories’ and ‘political
stability’ are more appreciated than profits by business leaders. The=
ir
class instinct tells them that lasting full employment is unsound from their
point of view, and that unemployment is an integral part of the
‘normal’ capitalist system Kalecki1943.
Uneven pre=
ssures
between workers and firms around this threshold of tolerance produced a
fragile, politically-driven dynamic induced by the “stop-and-go”
character of social expenditure. As the role of government in the determina=
tion
of economic activity increased, it eventually rouses the opposition of busi=
ness
leaders, as the new state of affairs strengthens the political undertow in
favor of workers.
Moreover, =
most
importantly, financial markets play a role via the state of economic
expectations. As the economy draws closer to full capacity, government acti=
on
becomes the rudder of economic activity, and crowds out the private
sector’s direct role in this regard. As the private sector’s ro=
le
decreases, financial markets will deem riskier the government debt used for
stabilization, align themselves with businesses given the increased clout of
workers and ask for higher yields. Higher borrowing costs signal policymake=
rs
to scale back spending, policy which deflates wages and renews business
competitiveness.
3
Stylized facts: The political content of fiscal policy.
As introdu=
ced by
Kalecki, these considerations apply to democratic market economies with
responsive governance structures with access to financial markets. Governme=
nt
borrows to shirk the recourse of higher taxes to pay for politically conven=
ient
expansionary activity. For internationally-linked economies, adjustment occ=
urs
mainly through the depreciation of the exchange rate, although domestically
such adjustment may also take place through an increase in government bond
yields. Under our model, adjustment takes place mainly through the latter, =
as
domestic financial markets will become jittery given lax thresholds of
political influence to popular opinion, as expressed in the desire to expand
deficits and increase the clout of domestic worker constituencies.
With respect to the former, [Perez-Caldentey(2009), Perez-Caldentey(2007)]
described in a series of papers the mechanism by which the above narrative =
held
for CARICOM countries. Using an analogous stock-flow model, he argued that
monetary circumstances provide hard constraints for growth in these countri=
es:
The
international financial architecture provides the framework for the working=
s of
‘real forces’. Indeed, were there no external constraints,
countries could pursue full employment policies through fiscal policy, or
proposals such as an international clearing union or a regional monetary
institution or regional fund Perez-Caldentey2009.
Indeed, wh=
ile
the papers do not delve on on the political dynamics affecting such fiscal
issues, he argues that fiscal reform faces perennial difficulties as extern=
al
shocks affect international demand for exports and governments in these
countries are pressured (and expected) to pick up demand. This frustrates s=
uch
calls for reforms as initiatives are captured and weakened by special inter=
est
groups in key strategic sectors.
Given these
conditions, external, government and often private sector deficits increased
the Caribbean stock of debt, and such debt burdens loop into difficult
refinancing choices. However, he wisely warned that “government
expenditures do not necessarily result in low growth or high debt levels. T=
he
outcome depends on the interaction
between government, external, and the private sector, an interaction th=
at
is the basis for stock-flow modelling [the emphasis is ours].”
[Perez-Caldentey(2009)].
A more gen=
eral
take on the interaction between these three sectors was proposed by
[Epstein(2001)] in his discussion about pro-labor and pro-rentier stances w=
hen
it came to central bank policy. In that paper, Epstein discussed how
financialization magnified the rentier motivations behind the relative
interests of industry and finance with respect to labor. He argued that such
policies depended on four factors: the productive structure of the economy,=
the
institutional structure of the central bank (i.e. its integration or
independence from government), the linkages between finance and industry and
the international position of the country.
For
[Kalecki(1943)], political tensions expressed endogenously in government
policy. For this reason, we capture both the exogenous and endogenous polit=
ical
pressures as noted with the wage, labor and financial yield target reaction
functions and exogenous government spending preferences. Taken together, th=
ese
lead to an examination of the institutional structure behind such spending =
and
the influence of financial markets, especially as institutional government
mechanisms constrain over the long-run the clout of worker voters. Barring
favorable borrowing terms (especially due to geopolitical considerations, l=
ike
those behind the exceptional privilege of the United States dollar (and the=
original sin of the rest of the
countries) [Eichengreen et al.(2002)Eichengreen, Hausmann, and Panizza]), no
government can borrow indefinitely to finance its budget deficit without a
paying political price. Indeed, reflecting on the motivations behind his
landmark 1979 paper, Anthony Thirlwall recognized that “there is a li=
mit
to the deficit/GDP ratio, and international debt/GDP ratio, beyond which
financial markets get nervous” [Thirlwall(2011)].
The use of=
debt
finances stop-and-go stabilization policy. This allows political groups to
balance broad electoral support and narrower business and financial interes=
ts
to keep incumbency. This is readily seen in Figure 1 below, which plots the
standard deviation of the primary balance and the average change in politic=
al
polarization as a proxy for class divisions for OECD countries. Increased
political polarization correlates with more volatile government budgets. Wi=
th
low taxes and high deficits, institutional compacts come under cyclical str=
ain
as majoritarian and elite interests alternate in power.
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image002.jpg")
Figure 1:
Scatterplo=
t of
the standard deviation of the primary budget balance (i.e. the difference
between current government spending and revenues from taxes) and the average
change in political polarization (defined as the probability that two deput=
ies
picked at random from among the opposition parties will be of different
parties) for 1980-2012 for selected OECD countries (Austria, Belgium, Denma=
rk,
Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, Norwa=
y,
Sweden, United Kingdom and the United States). Source: World Economic Outlo=
ok,
IMF October 2012 http://www.imf.org/external/pubs/ft/ weo/2012/02/weodata/index.aspx and World Bank 2012 Database of Politi=
cal
Institutions http://go.worldbank.org/2EAGGLRZ40.
4
The model
Businesses,
market structure and investment: Accrued profits to business with respe=
ct
to output is the profit share π=
;,
where π =3D (1 − ψ), where ψ is the wage share. Businesses and have pricing power over
market production through a markup =
τ
over average variable costs, comprised of labor costs. Relative bargain=
ing
power over workers determines the markup:
=
; &n=
bsp; =
; &n=
bsp; =
; ![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image004.gif")
&=
nbsp; &nbs=
p; &=
nbsp; &nbs=
p; (1)
Investment is a function of investment
demand as determined by the growth rate of capital stock ![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image006.gif")
and the capital growth rate, as all=
owed
by saving supply, ![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image008.gif")
.
Growth in capital expenditure demand is a function of the profit rate![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image010.gif")
(where K is the level of capital
stock, fixed at K¯ over the
short-run), the output-capital ratio ![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image012.gif")
and
autonomous investment demand I<=
sub>0 (or
animal spirits):
=
; &n=
bsp; =
; &n=
bsp; =
; gi =3D I0
+ [g1(1 −=
; ψ) + g2]u &=
nbsp; &nbs=
p; &=
nbsp; &nbs=
p; (2)
Where g1,g2 > 0. In turn, growth in savings is determined by the after-t=
ax
savings of workers sψ and
businesses sπ o=
ut of
output:
=
; &n=
bsp; =
; &n=
bsp; =
; gs =3D [sπ(1
− ψ) + sψψ](1 −=
; t)u &=
nbsp; &nbs=
p; &=
nbsp; &nbs=
p; (3)
Where sψ,sπ
> 0 and t is the t=
ax
rate. Business owner consumption is a portion b out of after-tax, unsaved non-invested profits and transfers,=
Cπ =3D b(1 − ψ)(1 − s
Workers:
The
wage share ψ is the total =
money
wage bill w with respect to out=
put Y which is the real wage per worke=
r ω divided by labor productivi=
ty![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image014.gif")
(i.e. where L is labor and Y is output).
=
; &n=
bsp; =
; &n=
bsp; =
; <=
/span>![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image016.gif")
&=
nbsp; &nbs=
p; &=
nbsp; &nbs=
p; =
(4)
Where P
are prices. Workers consume their after-tax, unsaved income plus transf=
ers
(of public goods, see below), so C<=
sub>ψ
=3D (1−sψ=
)(1−t)ψY
+pψ, where =
Cψ is worker
consumption, sψ is
worker saving and pψ is
a public good transfer to workers.
The government policy function: ![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image018.gif")
and
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image020.gif")
0
(where upper script denotes partial derivatives with respect to p).
This welfare function in spending
maximizes majority rule in total consumption of private and public goods of
business owners and workers subject to available tax resources over the
polity’s policy horizon T=
. The
welfare function Λ is:
T
&nbs=
p; &=
nbsp; &nbs=
p; &=
nbsp; &nbs=
p; Λ =3D=
X(Cψ + Cπ) &=
nbsp; &nbs=
p; &=
nbsp; &nbs=
p; &=
nbsp; (5)
t=3D0
Subject to the government constraint G =3D T
+ D(1 − i) − G0 − H=
(p) + R
where D is total government
(debt) borrowing over T, iD is interest payable on governme=
nt
debt and T are collected taxes.
Additionally, the constraint includes the democratic rents in securing such
majority rule by politicians, as denoted by R
(where R =3D R(ψ), and Rψ > 0). On that
note, the government policy function becomes:
&nbs=
p; &=
nbsp; &nbs=
p; &=
nbsp; Max Λ s.t
T + D − G0 − H(p)
− iD − R =
; &n=
bsp; =
; &n=
bsp; (6)
Using Lagrange multipliers λ in the function L:
&nbs=
p; &=
nbsp; &nbs=
p; T =
&nb=
sp; =
T
L =3D X(Cψ + C<=
sub>π)
+ λX(T + D
− G0 ͨ=
2; H(p)
− iD − R) =3D 0
&nbs=
p; &=
nbsp; &nbs=
p; t=3D0 &nb=
sp; =
t=3D0
=
; &n=
bsp; ![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image022.gif")
&=
nbsp; &nbs=
p; &=
nbsp; &nbs=
p; &=
nbsp; (7)
Solving fo=
r the steady
state costs of politicians to secure a majority under such a policy problem
leads to:
=
; &n=
bsp; =
; ![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image024.gif")
&=
nbsp; &nbs=
p; (8)
Under this policy problem, the costs f=
or
majority rule require higher relative consumption by workers of the marginal
public good as a share of all consumable available public goods, in additio=
n to
the marginal production of such goods for this social group.
Political influences in the government budget=
and
deficits: Given
this policy function, democratic governments would face politically unbalan=
ced
budgets over the planning horizon as ruling interests cater to labor groups=
to
keep in power. Rents required to secure a majority consistently imply a
political provision of spending beyond what is allowed by the budget
constraint. For this reason, some institutional mechanism implicit in
government expenditures must restrain political expenditures over the short=
run
periods as to make Rψ∗ as close to zero. Doing so requires
lowering the bargaining power of labor at the level of G where p and ψ are maximized (we call this=
G, where![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image026.gif")
To keep the government budget constrai=
nt,
government would need to endogenously restrict the provision of the public =
good
at G∗ to make Rψͨ=
7; − γ(ψ) =3D 0. The budget deficit D˙
is a function of autonomous government expenditures G0 and the provision of the public good g, collected taxes T, interest payable on government debt iGD and a political reaction function γ that itself is determined b=
y the
sensitivity of spending to the the wage share ψ:
=
; &n=
bsp; =
; &n=
bsp; =
; ∆D =3D G0
− T + iGD + [p − γ(ψ)] &nb=
sp; =
&nb=
sp; (9)
The political reaction function γ measures the elasticity of
government expenditures with respect to the wage share. Arguably, it is the
political sensitivity of government decreasing expenditures when the bargai=
ning
power of workers increases. Total tax collections are taxes paid by workers=
and
businesses of their respective claims to output (we assume a flat tax t =3D t¯over
wages and profits), hence T =3D=
(1
− t¯)(1 − ψ)Y + (1 − t&ma=
cr;)ψY =3D tY¯ .
Financial
market influence in fiscal policy: Government reacts to financial markets
given how the latter discount new offerings of sovereign debt, using intere=
st i as the running yield (and refina=
ncing
cost) of such (zero-coupon) government securities. Indeed, such yields desc=
ribe
how new and existing debt burdens are discounted, priced according to the d=
ebtto-output
ratio.
=
; &n=
bsp; =
; &n=
bsp; =
; &n=
bsp; ![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image028.gif")
&=
nbsp; &nbs=
p; &=
nbsp; &nbs=
p; &=
nbsp; (10)
Where φ1 > 0. With this in mind, we can marginally modify γ to internalize how governme=
nts
react to international bondholders given a debt-to-capacity target (as give=
n by
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image030.gif")
). Indeed,=
the
government’s reaction to debt markets depends on how politically feas=
ible
are interest and principal payments to be paid on outstanding liabilities o=
ver
the long term. Taken together, then Eq. (9) becomes:
=
; &n=
bsp; =
; ![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image032.gif")
&=
nbsp; &nbs=
p; (11)
Where government spending is ![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image034.gif")
and γ1,γ2 are the politi=
cal
sensitivity to worker bargaining power and debt levels respectively.
The
financial sector: The financial sector issues deposit accounts to accommoda=
te
households and business owner saving, offers loans to finance capital
expenditures for businesses and buys government securities. It pays a iS rate on worker Sψ and business ow=
ner Sπ savings and rec=
eives iL and iD from loans L and
government securities D respect=
ively.
On that note, the deposit-loan interest risk spread iL − i=
S
is defined by the running yield on government securities iG:
=
; &n=
bsp; =
; &n=
bsp; =
; &n=
bsp; iL − iS =3D i
Macroeconomic balance: To close t=
he
model and achieve macroeconomic balance, we simply state gi + gd=
−
gs =3D πB where πB
is the profit rate of the financial sector ![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image036.gif")
as determined by Equations 2, 3 and=
9
above and the transaction matrix below, and![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image038.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image040.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image042.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image044.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image046.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image048.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image050.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image052.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image054.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image056.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image058.jpg")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image060.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image062.gif")
![](3D"05EloyFisher-SFCPBCKaleckiRevisitedv1(1)_archivos/image064.jpg")