Transcript (I + A)/(I – A) -1 - Political Economy of transition

Economics of Transition
Lecturer: Prof. Hartmut Lehmann,
University of Bologna;
IZA, Bonn.
1. Economics of Planning – Historic
developments







Founding of the SU
War Communism
New Economic Policy
Industrialisation Debate
Industrialisation and Collectivisation under Stalin
“Mature” socialist system
Export of this system to those countries that had
been liberated from the Nazis by the Red Army
(CMEA area)
Aside

Discussion between Janos Kornai and Richard
Portes
Kornai: system is irrational at the macro level 
firms (and workers) behave irrationally at the micro
level
Portes: Firms and workers behave rationally at the
micro level, but are confronted with irrational
constraints at the macro level
Ten years of transition make one lean more towards
Kornai’s view – anyway we present here mainly
Kornai’s and Hare’s views of central planning
Coordination mechanisms in the society
(For understanding of central planning we need to discuss
these first)
There are 5 types of coordination:





Bureaucratic coordination
Market coordination
Self-governing coordination
Ethical coordination
Family coordination
Dimensions of coordination mechanisms

Structure of coordination

Types of flows of information, communication
enabling coordination

Incentives, motivation for participants
1. Bureaucratic coordination






Superiority-subordination relationship between the
coordinator and the coordinated (“vertical linkages”) 
There exists a hierarchy of superiority-subordination on
several levels
A superior of subordinates is a subordinate to a higher up
superior.
Vertical linkages are asymmetric – subordinate is much
more dependent on superior than vice versa.
Superior is not chosen by subordinates but appointed from
above.
Information mainly via command from superior to
subordinate.
Incentives to carry out command: subordinate wants
approval; receives award offered, avoids legal sanctions for
not compliance.
2. Market coordination




Sellers and buyers who have horizontal
linkages; both rank equal in legal terms.
Many forms of informational flows between the
two, most of which is price.
Voluntary exchange motivated by the
expectation of a material gain from transaction.
Coordination is normally monetised: good in
exchange for money.
3. Self-governing coordination






Lateral relations, i.e. equal members of selfgoverning association.
Coordination is conducted according to a constitution,
a book of rules.
Normally some vertical linkages, as members
choose committee to run association.
But (!) members can dismiss committee (unlike
type 1).
Informational flow: numerous kinds, most important
are votes and collective decision.
Motivation: above all awareness that interests of
member coincides with collective interest.
4. Ethical coordination





Mainly horizontal linkages – donor and recipient
Different from type 1 in that recipient is not
afraid of donor and different from type 2 in that
donor is not interested in material gain.
Informational flows takes mainly the form of
application (recipient) and offer (donor).
Donor’s motivations are altruistic arising from
political or religious beliefs.
Often money absent in transaction (unlike other
main horizontal linkage coordination – market
coordination).
5. Family coordination




Participants bound by family ties (types 1 –
4: participants usually are not bound by family
ties).
Vertical and horizontal linkages (depends
on the specific structure of a family)
Communications: all kinds.
Motivation: love and/or sense of duty
arising from customs.
Examples of various coordination types





(1) – army, police, internal administrative
apparatus of a large firm.
(2) – Market Hall, Bazaar, department store,
stock exchange
(3) Choir, Professional Association
(4) Oxfam
(5) Communal consumption in a household.
The Classical Socialist System and
Coordination Mechanisms
1. Bureaucratic coordination





This type dominates very strongly.
While under capitalism powerful organizations exist
that are bureaucratically coordinated, CSS for the first
time organizes virtually all social, economic, and political
activity along the lines of this mechanism.
Undivided political power, control of virtually all means
of production and bureaucratic coordination go hand in
hand and give CSS its specificity.
The power elite has exclusive right of disposal over means
of production, is hierarchically structured and tries to
coordinate whenever possible by bureaucratic means.
State-owned enterprises (SOE’s) are bureaucratically
managed, as to the last worker on the shop floor vertical
linkages dominate.
1. CSS and Bureaucratic coordination,
cont.



Under CSS we do not have a perfect hierarchy,
i.e. a subordinate has just one superior.
Normally a person’s activities are subject to a
layer of controls. Political police, inspection
committees in ministries, trade unions,
“Blockwart” control activities.
This seems dysfunctional, but is perfectly
rational since superiors do not trust
subordinates. Redundancy guarantees that if one
control mechanism does not work properly, e.g. 3
others will.
1. CSS and Hirschman’s “exit” and “voice”








“exit” and “voice” in a general economic context
“voice” exists, although only feebly: criticism of party
elite, political and economic system per se is not allowed
(minor elements of system may be criticized). Dangerous
to criticize superiors (the higher up they are, the more
dangerous)
Upshot: “voice” in most cases below a critical threshold
that is necessary to improve the functioning of the system.
“Exit” virtually impossible, e.g.
Leaving party gives a bad signal
Leaving firm and then what? I can only go to other stateowned firm.
Leaving the country for another system is forbidden.
So, because “exit” is virtually non-existent, voice is so
weak (when “exit” is an option, organizations will allow
more voice to prevent “exit”.)
2. CSS and Market Coordination
Supplying
sector
SOE’s as
consuming
sector
Coops as
consuming
sector
FPS as
consuming
sector
IFPS as
consuming
sector
Households as
buyers of goods
and services
SOE’s
B
B
B+M
0
B+M
Coops
B
B
B+M
0
B+M
Formal
private
0
0
M
(w/ B)
M
M (w/B)
Informal
private
sector
0
0
M
M
M
Households
as sellers of
labour
B+M
B+M
0
M
-
Allocation of
investment
resources
B
B
0
0
-
sector
2. CSS and Market Coordination, cont.




Entry (1,1), i.e. flows between SOE’s will be
discussed in detail under planning.
(1,1) is for most part bureaucratic, there are
some residues of market coordination
(“tolkach”). (1,2), (2,1), (2,2) same as (1,1),
i.e. no difference between SOE’s and
cooperatives
FPS ((1,3),(2,3) allowed to buy inputs from
SOE’s and Coops, regulated by bureaucratic
rules and at higher prices (market price
higher than official price
Normally, FPS not allowed to sell its product
to SOE’s and Coops (3.1, 3.2 – 0 entries)
FPS interacting with itself and with households.
Market but subject to some bureaucratic
regulations.
Pure market only when informal sector is
involved (3.4, 4.3, 4.4, 4.5)
IFP’s cannot supply to SOE’s and Coops.
1.5, 2.5 – …. plus commercial channels (heavily
regulated)
5.1, 5.2 – combination of market mechanisms
and bureaucratic coordination
5.3=0 as in CSS FPS are not allowed to have
outside labour.
Labour allocated partially through
market coordination, capital only
through bureaucratic coordination (6.1,
6.2 = B)
FPS cannot receive capital from the state.
State owned sectors B
Privately owned sectors M
Self-governing coordination
Initial phases of socialism were not only on
paper, however, it was a fiction under CSS (1
dominates 3)
Proof: “Elected” officials are chosen from top
and are not accountable to voters.
Ethical coordination
Initially enthusiasm generated many
activities of this coordination kind.
Degenerated soon even already under
Lenin.
Family coordination
Since most of the countries where CSS
ruled were underdeveloped, this type
of coordination was rather important.
Initially (with land reform) became
even more important.
Finally, though, forced industrialization
and forced growth policies resulted
in large damage to this coordination
mechanism
Planning
Large literature on this prior to collapse of the
Communist block.
One strand of literature:
Given perfectly competitive general equilibrium as a
benchmark, can a central planner (in principle)
reach the same type of (also in the West)
hypothetical equilibrium where all markets clear,
i.e. where supply=demand in each product group?
Given assumptions: no asymptotic information, no
externalities.
With firms maximizing π, consumers maximizing their
utility, prices carry essentially enough information to
adjust behaviour of firms in such a way that S=D
Planning contd.
In the absence of underlying behavioural motivation by
firms and consumers, state needs to step in ,
determine the overall level of “zillion” of goods
produced by firms and consumed by firms, the state,
and consumers.
Input-output techniques are used to achieve this.
While in principle this might be done achieving roughly
overall balance, 2 points can be made.
1) It might be quite difficult to get balance for each
good, which implies (2).
2) Consumer is treated as a “residual agent” in this
system: first have the demands of firms and the
state be fulfilled; if anything is left over, it is given to
consumers.
Planning contd.
Leading up to a general point:
The socialist system was initially quite successful at
the macro level:
1) High growth rates
2) Full employment
3) Seemingly no inflation
The problem, also with the reformed versions of the
CSS (Hungary and Poland) is at the micro level:
firms do not maximize π and consumers certainly
do not maximize their utility => a very wrong
incentive structure which leads to gross
inefficiencies in the use of factors of production and
eventually to stagnation and decline (i.e., first
Δy >>0; then Δy ~= 0, finally Δy <0)
To see this, let’s look at policy
instruments used in the SU
Policy object
Enterprise activities
Prices, taxes, and
subsidies
Dealt with under
Central planning
Inputs, outputs
regulated by direct
instructions from
above
Mainly fixed centrally.
Wide variety of
taxes and subsidies
Policy instruments used in the SU
Policy object
Profits
Wages
Dealt with under
Central planning
Flow to the state budget
after enterprise funds
have been formed
Central wage scales
(rather compressed),
wage fund controls at
the enterprise level
Policy instruments used in the SU
Policy object Dealt with under Central
planning
Controls related to the degree
Credit
of plan fulfillment of each
enterprise
Investment Apart from very small projects,
all projects need to be
approved at levels above the
enterprise level. Often in form
of non-repayable budget
grants
Policy instruments used in the SU
Policy object
Dealt with under Central
planning
Budget
Always aims for balance
(macroeconomic
fiscal policy)
Foreign trade
Wide variety of taxes,
subsidies and direct
controls. State
monopoly
Very weak spot: foreign trade or in
general relations to the outside world.
[ These relations are important if
centrally planned economy needs
modern investment goods, but has no
control over world market prices and
over import prices either ]
[ Clearly, the CSS is not very agile and
cannot react rapidly to changing
exogenous prices ]  see also Hayek
Taxonomy (various dimensions one could
use to characterize a CSS but also to compare
it to other types of planning systems, e.g.
France before 80’s and 90’s)
(1) Periodization of plans

Short-term (one year or less) – in
operational sense the most important one

Medium-term (about 5 years) – politically
most important

Long-term (15-20 years) – pie in the sky
[ Note: often with change in regime medium-term
plans are severely revised and altered ]
Taxonomy contd
(2) Detail and scope of plans: for single
enterprises, for sectors, and
economywide; all three present in one
plan.
(3) Planning organization: number of
levels in planning hierarchy and
relationship between them, functional,
branch, and regional bodies involved.
(4) Information flows. Vertical and
horizontal flows, economic content of
major information flows.
Taxonomy contd
(5) Policy instruments. Main distinction:
administrative instruments and price-type
instruments. In CSS first prevail.
(6) Incentive systems. To elicit reliable
information, precise plan-fulfillment or
good overall performance. Incentives can
be implicit in price and tax system or
explicit in the system of plan indicators.
(7) Preferences. Planners’ preferences and
consumers’ preferences. In CSS first
decisively prevail.
(8) Tautness. Explain concept of tautness.
The mechanisms of planning
(1) Construction of the plan
(2) Implementation of the plan
Point (1)
a. Organizational framework
Organized according to B, nearly
everything in the construction of a
plan is vertical.
CPB
Ministry 1
Ministry 2
Chief
Chief
Administration 1
Administration 2
Enterprise 1
Enterprise 2
Functional agencies
………….
………….
Ministry m
Chief
Administration k
Enterprise n
The mechanisms of planning ctd
Situation gets a bit more complicated as
we have planning on the
- SU level
- regional (republic) level
- union-republican level
Council of Ministers = senior executive
council of the Soviet government
(important caveat: party apparatus
parallel machine similarly hierarchically
structured like Soviet government and
actually making the basic strategic
decisions)
The mechanisms of planning ctd
Attached to the Council is Gosplan – State
planning commission (Государственный
комитет планирования – Госплан)
responsible for actual detailed planning at
the national level.
Ministries (covering major industries – about
35 in 1977) – subordinate to the Council.
Ministries have departments called “Chief
administrations” (later called “Industrial
Associations”)
Some administrations are responsible for parts
of a ministry productions, others – for a
specific function, e.g. finance.
The mechanisms of planning ctd
At the bottom – enterprises, some of which
have been amalgamated into larger units
(“production associations”).
There were 44 000 basic units in Soviet
industry in 1977.
They are subordinate to a Ministry via a
Chief Administration.
This is the main hierarchy for planning the
production
and
controlling
the
implementation of the plan.
The mechanisms of planning ctd
There are other important organizations:
(1) State Committee for Supply (Council of
Ministers) – controls supply of inputs to
enterprises.
(2) State Bank, both Central bank and
monopoly commercial bank; supervises
financial transactions authorized by plan.
(3) State committee on prices (Council of
Ministers) sets wholesale and retail prices.
(4) State committee on labour and wages –
sets wages and salary levels.
(5) State
committee
for
science
and
technology
Complex hierarchy where information
travels up and down. Overloaded with
information. Information is often disturbed
when it comes back to enterprise level.
Chief Administration
Info
Info
Enterprise 2
Enterprise 1
Info
b. Construction of a plan
15-20 –year plan: some forecasting
exists which is rather meaningless
5-year plan: its main economic
function is to develop an investment
plan that makes fulfillment of 1-year
plans possible and that makes
consecutive
one-year
plans
consistent with each other
Also
some
political
functions:
“mobilizes the masses” propaganda,
piece at the party congress
1-year plan: often partitioned into quarters, is by
far the most important plan among the 3 types.
This plan sets out targets of output etc. for all
enterprises and is legally binding (criminal
sanctions can in principle be applied against a
manager for not implementing the plan).
Construction and implementation of a plan are
interrelated as implementation is controlled with
the constructed plan as a benchmark.
Bonuses for managers for fulfilling the plan
Managers try to influence the construction of
the plan (to have a modest plan).
Time frame:
Plan for production in the year 2:
year 1
year 2
Planning year
Planned year
Stages of construction
(1) Council of Ministers (guided by the Party)
sets targets of growth rates for major
sectors of economy (those deemed the most
important, not necessarily the largest ones)
(2) Gosplan receives these targets => derives
a set of output targets by major product
groups. Targets are set knowing or evaluating
output capacity of each industry. These
targets then are sent to ministries.
(3) Ministries further disaggregate targets received
from Gosplan Chief Administrations and output
targets for particular product groups  enterprises
receive output targets (stage completed by April of
the planning year).
(4) Enterprises calculate their input requirements 
regional supply organization aggregates all input
requirements  national supply organization 
Gosplan
At the same time enterprises negotiate with Chief
administrations for revised output targets –
production targets within industry may be reallocated
(without (!) changing statement of input needs).
(5) Remainder of the year is used to ensure
consistency between production and other sources
of supply AND input requirement and other sources
of demand.
The tool used for that is the “material
balance”.
Gosplan calculates balances for the most
important product groups (~2000) ~ 80% of
the value of production.
Supply side
Demand side
(1) Current production
(2) Imports
(3) Stocks brought
forward
(4) Intermediate inputs
(5) Consumption
(6) Investment
(7) Exports
(8) Stocks required at
the end of the year
Total required
Total available
Entries are done by May of the planning year
Then Gosplan can determine which products
are in deficit (excess demand) and which in
surplus (excess supply).
Historically often taut planning => many goods
in excess demand
Within each balance planner can either reduce
demand or increase supply if ED
(6) and (7) cannot be reduced (growth is
reduced or balance of payment changes
unfavorably.
(8) – tiny item, (5) is historically residual but not
as time progresses. Reducing demand is
difficult. Raising current production is the
only viable route
Problem with this: it increases input
requirements, i.e. output of other sectors
needs to be raised => raising second
round of input requirements. If there was
enough time, the new “disequilibria”
could all be taken into account, but there
is not enough time => juggling the figures
(6) End of the planning year, plan gets the
final approval of the Council of Ministers 
enterprises (via ministries and Chief
administrations).
Supply organization prepares attachment
plan assigning inputs to customers.
Supplier and customer complete contracts
with supply provisions now very detailed.
(1st time that 2 enterprises have official
contact)
One more issue – that of aggregation
of different products.
1) Prices are used above all for that, i.e.
prices have very little allocative
function.
2) Tons, square meters, units of
homogeneous output.
[ Enterprise level output target is in
terms of gross value of the product
+ specific products in physical
units ]
I-O-Analysis of 2-sector economy
yi = final (net) output of product i (i=1,2)
xi = gross output of product i, including
intermediate inputs (i=1,2)
aij = input of product i per unit of output of
product j (i,j = 1,2) – input coefficients
x1 = a11x1 + a12x2 + y1
x2 = a21x1 + a22x2 + y2
Gross output = Intermediate demand +
Final demand
Solving for x1 and x2
(1-a22)y1 + a12y2
x1 = --------------------------(1-a11)(1-a22) – a12a21
a21y1 + (1-a11)y2
x2 = -------------------------(1-a11)(1-a22) – a12a21
Given the aij’s (technological input-output
coefficients) for any level of final demand (y1,
y2) we can calculate gross output (x1, x2)
necessary to sustain it.
In matrix notation
x = Ax + y => x – Ax = y
=> x(I – A) = y
=> x = (I – A)-1y
If elements of A are direct input coefficients (e.g.
amount of coal required to make 1 ton of
steel), elements of (I – A)-1 mean total input
coefficients, both direct and indirect (e.g.
amount of coal required to produce 1 ton of
steel, amount of coal required to make other
inputs into steel, amount of coal required to
make inputs for those inputs etc.)
A – matrix of direct input coefficients
(I – A)-1 – matrix of total input coefficients
(I – A)-1 computed by inversion; can also be
computed by iteration
E.g. one-sector economy
x1 = a11x1 + y1,
x1 – a11x1 = y1
x1 = (1 – a11)-1y1
Iteratively:
To produce one unit of net output, gross output
must comprise: net output + direct input into
net output + input into that direct input + input
into that first-round indirect input + …
x1 = (1 + a11 + a112 + a113 + …)y1
From algebra we know that if a11<1, a
sequence converges to
1/(1 – a11) = (1 – a11)-1
x1 = (1 – a11) -1y1
Multi sector:
(I – A) -1 = (I + A + A2 + A3 + …)
So x = (I + A + A2 + A3 + …)y
Related to Material Balances approach
If I-O approach would be used in planning,
we would start with y and then try to find
x that is consistent.
Instead what is done is to have an initial x0
and y, then by iteration a consistent x
(vector of gross output target) is
computed (y might also be adjusted
during this).
1st iteration: communication initial output
targets xo to producers  get input
requirements Ax0.
Subsequent iterations: in the planning
offices done balance by balance =>
repercussions for whole economy, not
just for sector in question
Max 3 rounds:
(I – A)-1 is substituted by
(I + A + A2 + A3 + …)
x1 = Axo + y
x2 = Ax1 + y = (I + A)y + A2x0
{A(Ax0 + y) + y = (I + A)y + A2x0 }
x3 = (I + A + A2)y + A3xo
x4 = (I + A + A2 +A3)y + A4xo
………………………
xn = (I + A + A2 +…+ An-1)y + Anxo
lim => x = (I – A)-1y + 0
n
If we know A, we can compute the amount
of inconsistency by ratio
(I + A + A2 + A3)
The smaller the
---------------------ratio, the more
(I – A) -1
inconsistency we have
For the SU: (I + A)/(I – A)-1 on average
50,6%
(I + A + A2)/(I – A)-1 on average 75,5%
(I + A + A2 + A3)/(I – A)-1 on average
87,8%
Implementation of a plan
The main point: plan is not
implementable
2 reasons:
1) Assumptions underlying the plan are
not fulfilled (random shocks and
outdated information)
2) Even on its own assumptions plan is
inconsistent the way it is constructed –
material balances in the end are closed
artificially => plan is not feasible
Consequences of an infeasible plan
a) Note: even if just one sector is
not able to fulfill output
requirements=> important
repercussions for the entire
economy as a whole
b) Late arrivals => “storming” =>
bad quality
Mitigating factors
1) Systematic ways to conceal production
possibilities and exaggerate input
requirements => buffer between one
sector and another sector.
Problem: severe underutilization of
resources
2) Enterprise tries to produce vital inputs
itself
Problem: poor quality of inputs => poor
quality of product; wrong specialization
of the enterprise
Mitigating factors contd
3) Unofficial system of supply
exchange (tolkach - «толкач»,
“fixer”) – guy responsible for illegally
procuring supply (quite risky)
4) Official measures (better channels)
to remove supply bottlenecks etc.
Microeconomic inefficiencies
Unwanted, unsold goods
Physical targets => gross allocative
distortions (“one meter” screws
instead of screws people need)
Allocation of labour and
consumption goods
LD planned => “Labour balance”
disaggregated by region and skill
Imbalances are rectified by economizing
or substituting types of labour.
Enterprises behave with labour as with
other inputs: exaggerate requirements
and have no incentive to economize.
Wage rates are administratively set,
differentiation by skill, sector of the economy,
sometimes region (also age)
Because of taut planning arises shortage of
labour => excess demand for labour => some
wage drift. {explain in western context}
LS: a function of wage rates, virtually never an
attempt to allocate labour administratively;
relatively high labour turnover in CSS.
Special topic: Hidden unemployment or “Labour
hoarding”- in western context;
In CSS: Storming and labour hoarding (given high
labour turnover and a shortage of labour)
 After liberalization one source of open
unemployment.
Supply of consumer goods
Market research to meet consumers’
demand at a disaggregated level => not
very successful
Consumer good is a residual
category?
Economic evidence:
see whether deviations of consumption
from its time trend are proportionately
stronger than deviations of total output
from its trend
C
Time trend
Actual
supply of
goods
Time
Consumption goods
Y
Time trend
Actual
output
Time
Total output
Portes and Winter
Evidence is mixed:
“Crowding out” of consumption goods for GDR
and Poland, reverse for CSR, Hungary.
SU - no crowding out found
--------------------------------------------L = level of employment
w = average wage
T = income taxes net of transfers
S –estimates of savings
Then nominal value of consumers’ demand
will satisfy equation: wL – T – S = pC
Central planners control all variables
apart from S, so they should be able to
balance income = value of demand
But in reality …
Retail P = Pw (1 + τ)
τ – turnover tax
In principle central planner could adjust τ
to equilibrate consumer goods markets.
i.e. D > S  τ
D<S↓τ
This is however not done in practice
Topic: repressed inflation
P
S
Economy is
supply
constrained
Pa
D (Y)
Q
=>↑ Money balances held by household
and/or ↓ Ls => shortage
While econometric models estimated in
70’s and 80’s show no disequilibrium,
there was, however, casual evidence of
queues and shortages + ↑ cash
balances in the FSU (tripling between
70 and 78) – “monetary overhang”.
Real test has been of course, when
liberalization occurred => corrective
inflation
P
P
S
S
D
D
Q
Industry more
competitive
Q
Industry more
monopolistic