Transcript Trade in Tasks and the Organization of Firms - uni

Trade in Tasks and the Organization of Firms
Dalia Marin, Jan Schymik, Alexander Tarasov
University of Munich
CESifo Global Economy Conference
Munich May 2014
The Changing Nature
of Organizations and International Trade
• Shifting international boundaries of the firm
trade in tasks or offshoring (Hummels et al 2001,
Marin 2006)
• Rise in CEO Pay in industrialized countries
• Move to decentralized management in corporations
(Rajan and Wulf (2006), Marin (2008), Marin and
Verdier (2010), Bloom, Sadun, van Reenen (2010)
Question
What is the connection between
trade in tasks,
decentralized management and
the rise in CEO pay?
Literature on Trade and Organizations
• Marin and Verdier (MV) (2008, 2010),
Caliendo and Rossi-Hansberg (2012):
North-North-trade in final goods leads to more
decentralized management
• Marin and Verdier (2012):
North-South trade in final goods leads to a ‚war for
talent‘ and to the emergence of the talent firm
• Grossman and Rossi-Hansberg (GRH) (2008):
Trade in tasks (offshoring) may lead to a rise in
wages in rich countries
Contribution of this paper
We incorporate trade in tasks of GRH (2008)
into a
small open economy version of the theory of
firm organization of MV (2012)
to examine
how offshoring affects decentralized
management and CEO pay in firms
By doing so…
we show that offshoring of production tasks
• leads to decentralized management in firms
in an open economy
• leads to improved competitiveness in
Northern firms as the productivity gains from
offshoring translate into gains in market shares
By doing so …
we show that offshoring of managerial tasks
• relaxes the constraint on managers (the
labor market effect) lowering CEO wages
• but toughens competition (the war for talent
effect) increasing CEO wages.
• leads to an unambiguous rise in CEO wages
in sufficiently open economies
Firm Survey of Offshoring Firms
We conducted a firms survey
among of 660 multinational firms in
Austria and Germany with 2200 affiliates
in Eastern Europe during 1990 – 2001.
80 percent of German investments in EE
100 percent of Austrian investments in EE
The Model
A Small Open Economy
Two sectors:
X-sector, differentiated goods,
monopolistic competition
Y-sector, homogenous goods,
perfect competition
Two factors of production:
human capital (skilled managers)
labor (production workers)
In the X-sector firms choose a firm organization
Three Firm Organizations
P-Organization:
P has formal authority in the firm
P runs the firm with A’s cooperation
A-Organization:
P delegates formal authority to A
A runs the firm with P’s cooperation
O-Organization:
P runs the firm without A’s cooperation
Firm Organization
We consider a firm with a simple hierarchy
CEO (P) hires a skilled manager (A) to run the
firm and workers to produce, P and A look for
projects for the firm
There is a conflict of interest between P and A
Payoffs of P and A depend on who’s project is
implemented
Trade-Off between Control and Initiative
P supervises the more
the larger are her stakes (the larger are
profits)
A has more initiative
the lower P’s probability of intervention
Cost of hierarchies: loss of initiative
The Choice of Decentralized Management
From Marin and Verdier (2012) we get
Firms will choose decentralized management
(A-organization) at an intermediate level of
profits
At low and high profits there is no trade-off
between control and initiative, hence, firms
choose control (centralized management)
Next step:
Endogenize profits:
small open economy with
monopolistic competition
Product Market and Trade Environment
Consumer preferences over the two goods
X and Y are
with
Ω, Ωm set of domestic and foreign varieties
Product Market and Trade Environment
In a small open economy (SOE): the number and prices
of foreign varieties are given, foreign demand for
domestic varieties is exogenous
Prices in the SMO are
n* number of foreign varieties, pm price of imported variety
n*(pm)1-σ = IM level of import penetration, exogenous
Offshoring of ProductionTasks
Production is sector X involves a continuum of tasks of measure
1. Performing each task requires c(i) units of labor.
It is profitable to offshore task j, j(0,1) if
, t(j) cost of offshoring
Marginal costs of the firm:
with the productivity gains from offshoring
ZX with IX the number of tasks offshored. IX = 0  ZX = 1
Profits of a firm i B
cB
Factor Markets
Factor demand depends on the equilibrium organization
P,A,O and on the organizational mix of firms.
k=P for low profits
k=A for intermediate profits
k=O for high profits
Equilibrium
Intuition
Upward sloping HH curve:
B/w increases, excess demand for managers, q/w has to
rise
q/w increases, few firms are looking for a manager,
excess supply of managers, number of firms n has to
rise which happens when B/w increases
Upward sloping EE curve (war for talent curve):
As B/w rises, firm entry, number of firms is fixed by H,
competition for managers, q/w rises
How are changes in openness affecting the
equilibrium?
Figure 2 An increase in openness
How are changes in
offshoring of production tasks affecting
the level of decentralized management?
Offshoring of Production Tasks
Intuition
Two opposing effects:
Zx   wc(i)/Zx   profits rise  (productivity
effect)
Other domestic firms become more productive as
well, revenues  and profits  (revenue effect)
In an open economy the productivity effect
dominates the revenue effect and B/w rises
Effect on Organization
When profits rise, P starts to monitor more
potentially destroying A’s initiative,
When profits rise sufficiently, the trade-off
between control and initiative favors
initiative, P delegates authority to the skilled
manager, decentralized management
Empirics
Measuring Decentralized Management
‘Who decides in your company over the
following decisions…..?
Please rank between 1 … 5’
1 taken at headquarters (CEO)
5 taken by the divisional manager
Level of Decentralization of Corporate Decisions
hiring a secretary
firing of personnel
hiring two workers
moderate wage increase
price increase of product
decision over product price
change of supplier
budget
introduction of new products
R&D expenditures
hiring >10% of current personnel
transfer prices
find acquisition
new strategy
financial decisions
decision over acquisitions
Austrian
Corporations
German
Corporations
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
Measuring Offshoring of Production Tasks
intra-firm imports from all subsidiaries in
percent of parent sales
instrument for offshoring of production tasks:
‘standardized foreign input’
Prediction
Prediction 1:
In a cross section of firms in an economy
open to trade, multinational firms will have
more decentralized management, when
they are offshoring more production tasks
to low wage countries.
Economic Magnitude
Offshoring firms are 33.4 percent
more decentralized
than non-offshoring firms
when we instrument offshoring with
‘standardized foreign input’.
Offshoring of Managerial Tasks
Managers perform a continuum of tasks of measure 1
requiring one unit of managerial labor.
Foreign managers do not receive authority in the firm.
Fraction of managerial tasks offshored is exogenous and
given by IS.
It is profitable to offshore managerial tasks if
q > q*
q, q* manager wages at home and abroad
q(1- IS) + q* IS cost of market entry
Offshoring of Managerial Tasks
Proposition 3:
In the P-equilibrium, there exists a cutoff level of
openness of the economy denoted by IM, such
that for IM > IMp: B/w and q/w are increasing in
Is;
and for IM < IMp: B/w is declining in Is, while
the impact of Is on q/w is ambiguous.
Offshoring of Managerial Tasks: IM > IMP
Offshoring of Managerial Tasks: IM < IMP
Intuition
Three effects of a rise in Is:
war for talent effect: Is  cost of market
entry   firm entry, fix H  q/w  and B/w 
labor market effect: Is   demand for skilled
managers   q/w 
competition effect: more firms find a manager
 n   competition   B/w 
Overall Effect on B/w
depends on the level of openness IM
large openness IM > IMp:
positive war for talent effect dominates the
competition effect, profits rise with Is
small openness IM < IMp:
competition effect dominates the war for
talent effect, profits decline with Is
Overall effect on q/w
depends on the level of openness:
large openness IM > IMp:
war for talent effect pushing up q/w
dominates the labor demand effect pushing
down q/w, as a result q/w 
small openness IM < IMp:
labor demand effect large (large shift in
HH) but war for talent effect also large
(steeper HH), effect on q/w 
Empirics
Measuring Offshoring of Managerial Tasks
‘How many managers of your parent company have
been sent to the affiliate firm?’
offshoring of managerial tasks = 1 – managers sent
in 57 percent of all direct investments at least one
managers has been offshored
on average 2.63 managers are offshored per
investment project
max 39 managers offshored
Measuring q/w
Total compensation per board member
of the firm (Kienbaum management
consultancy) relative to average wage of
the firm (firm survey)
Prediction
Prediction 2:
In a cross section of firms in sectors
sufficiently open to trade, multinational
firms will have more decentralized
management when they are offshoring
managerial tasks to low wage countries.
Economic Magnitude
An increase in the sample mean of the
fraction of managers offshored (1.48)
reduces the level of decentralized
management by 3.1 percent but increases
the level of decentralized management
by 4 percent in industries with a level of
openness above the 25th percentile of the
openness distribution.
Prediction
Prediction 3:
In a cross section of firms exposed to
international trade, multinational firms will
pay their CEOs lower wages when they
are offshoring managerial tasks to low
wage countries and they will pay their
CEOs higher wages when the number of
firms in the domestic market increases.
Economic Magnitude
One additional manager offshored lowers CEO
pay relative to workers by 6.9 percent (labor
market effect).
This implies that relative CEO compensation is
lower by 13 – 18 percent due to managerial
offshoring.
One additional manager offshored allows one
additional firm to enter which increases relative
CEO pay by 2 percent (war for talent effect).