Download GE05-Riezmann-06DEC 225733 en

CESifo Area Conference on
Global Economy
10 – 11 December 2004
CESifo Conference Centre, Munich
Globalization Distance Metrics
for the World Economy
Ray Riezman, John Whalley
and Shunming Zhang
CESifo
Poschingerstr. 5, 81679 Munich, Germany
Phone: +49 (89) 9224-1410 - Fax: +49 (89) 9224-1409
E-mail: [email protected]
Internet: http://www.cesifo.de
Globalization Distance Metrics For The World Economy ∗
Raymond Riezman
Department of Economics, University of Iowa
Center for Economic Studies - Institute for Economic Research (CES-ifo, Munich)
John Whalley
Department of Economics, The University of Western Ontario
National Bureau of Economic Research, Inc. (NBER)
Center for Economic Studies - Institute for Economic Research (CES-ifo, Munich)
The Centre for International Governance Innovation (CIGI, Waterloo)
Shunming Zhang
School of Business, Nanjing University
Department of Economics, The University of Western Ontario
December 2, 2004
Abstract
An earlier paper of ours discusses metrics of globalization between fully integrated and
trade restricted equilibria for individual price taking economies. These reflect the distance
from full global integration due to trade barriers, barriers to factor flows, barriers to international financial intermediation, slowed technological diffusion and other economy specific
features yielding less than full integration into the global economy. Here we discuss distance
measures from integrated free trade for the whole of the global economy. Important differences arise relative to single economy metrics since globalization measures for the world can
be small if small economies are integrating into large economies, rather than all economies
treated as separating integrating into a larger world. Importantly globalization distance
measures for individual economies are also shown to depend on both domestic and foreign
barriers to goods and factor flows.
∗
This is a draft of a paper to be presented at a CES - ifo conference to be held in Munich, Germany, December
10 - 11, 2004. The second author acknowledges financial support from SSHRC (Ottawa).
1
1
Introduction
An earlier joint paper of ours (Riezman, Whalley and Zhang (2004)) discussed how close
particular economies are to an equilibrium characterized by full integration into the global
economy (such as free trade in goods, full factor mobility, or both) where each economy was
treated as a small open price taking economy and we assumed such equilibria are unique. The
motivation we offered was that there was an emerging popular literature in the Economist,
Foreign Policy and other media outlets reporting indices of globalization by country, but little
or no research literature on appropriate metrics. We suggested that what was at issue was an
explicit comparison between equilibria using distance metrics 1 , and we also highlighted the
main focus of prior general equilibrium literature as being on comparative statics and issues of
existence, uniqueness, and stability (see Arrow and Hahn (1971), and Mas-Colell (1985)) not
measures of distance between equilibria.
We formalized alternative metrics of globalization as distance measures between equilibria,
and explored the behaviour of these metrics in the single price taking economy case using
data to yield simple forms of integration restricted equilibria for some sample economies. We
used data for ten OECD countries (Australia, Canada, Germany, Italy, Japan, Korea, Mexico,
Norway, UK, US) and made calculations of globalization metrics for each using these measures.
These were chosen as a sample of countries varying by size, level of income per capita, trade
pattern, and size of trade barriers. Our conclusion was that globalization distance metrics can
behave in different ways and numerical measures of the degree of globalization for individual
economies can therefore be hard to interpret. No unambiguously preferred metric seemed to
offer itself despite the growing importance attached to distance metrics in globalization debate.
In this paper our focus is instead globalization measures on the whole of the global economy
rather than on single economy measures, and we ask how far the world is from complete
globalization or integration (or integration). We first show how measures similar to those
for individual economies can be constructed using a model a the global economy taken as a
single entity. Some yield a single index measure, others measures for individual economies. To
1
Measures of distance between equilibria are also critical in a number of other current subareas of economics.
In the calibration area, for instance, inexact calibration (see Dawkins, Srinivasan, and Whalley (2001)) involves
choosing parameter values for equilibrium structures so as to produce model generated equilibria as close as
possible to observed data (pre-adjusted for compatibility with model equilibrium conditions), and closely related
metrics of distance between equilibria are also needed here.
2
implement certain of the measures we propose we require some form of global social welfare
function, which we also discuss in the text.
We then discuss how this approach to globalization metrics affects perceptions of the extent
of globalization, compared to our earlier approach. For large economies, with endogenous global
prices as trade and other barriers removed or modified, little change in domestic prices need
occur. In effect, small economies will now integrate into larger economies and globalization
distances for large economies may be small even if their own barriers are large. Hence globalization metrics constructed for individual economies can be misleading if their relationship with
the wider global economy is not also considered. Any computed variance in separately constructed measures across individual economies conveys little about the closeness of the world to
full integration. Also, barriers in foreign markets influence the globalization metric for a given
economy, as well as (and in some cases more so than) barriers employed at home.
Our conclusion is that this new round of analysis of globalization metrics casts further
doubts over the interpretation of measures of globalization that percolate the public domain
beyond those raised by our earlier paper.
3
2
Globalization Distance Measures in A Global Economy
In this section we present some analytics for distances measure which can be used to assess
how close to or far away an equilibrium for the world economy is from that which would characterize full integration by all economies into the global economy (globalization). We assume
that we are to compare a global trade (or factor flow) restricted equilibrium to a full integration
equilibrium and only one of these equilibria will be observed (typically a global equilibrium in
the presence of trade and factor flow restrictions). We first calibrate a model of global trade
to data in the presence of such restrictions, and then use the parametrization generated for
this model to compute an unobservable free trade global equilibrium. Globalization measures
involve some form of pairwise equilibrium comparison. We make the strong assumption that
both free trade and restricted equilibria are unique 2 . These measures are closely related to
those for individual economies presented in Riezman, Whalley and Zhang (2004).
We begin with simple measures of distance between free trade and trade restricted global
equilibria calculated by summing the squares of differences across equilibria in endogenous
variables (prices, quantities). As Riezman, Whalley and Zhang (2004) note there are several
difficulties which arise with such measures. One is that if price variables are involved they are
not invariant to alternative price normalizations. Another is that the rationale for including
all variables in such measures (such as both prices and quantities) is not clear, while neither
is it clear that some variables should be excluded. If only a subset of variables are included
in such distance measures one has to rationalize which they are and why they should be so
used. Such metrics could also involve exogenous variables such as endowments, although for
simplicity these are not used here.
A second type of distance measure we use involves computing excess demands in the global
economy at non-equilibrium prices. For these measures we take the prices from one equilibrium
and use them in the model of the global economy generating the other equilibrium and calculate
the size of the implied global excess demands. Thus, if we calibrate our global model using
data generated in the presence of tariffs, we can introduce the free trade equilibrium prices
into the associated model parameterzation and compute global excess demands (i.e. the sum of
country imports and exports). We take the absolute value of the excess demands generated in
this way relative to total global demands as the distance measure between the two equilibria.
2
But see the discussion of the likelihood of multiplicity of equilibria in models similar to those we use in Kehoe
(1991) and Whalley and Zhang (2004)
4
These are model dependent measures in that the numerical value of the distance measure will
vary as the underlying model parametrization used to calibrate to the equilibrium data in the
presence of country tariffs changes (say, as elasticities of substitution and share parameters in
CES functions change).
These are also problems with these types of measure and these are similar to those identified
by Riezman, Whalley and Zhang (2004) in the small open economy case. One is that such
measures are only operational in the sense of producing a single value where there are point-topoint mappings (a pure exchange economy), not correspondences (economies with production).
Another is that one can have pairs of equilibria for this class of measures which yield sharp
differences in distance measures (close, far) in prices and quantities.
Finally, we use a third class of global distance measures constructed in model parameter
space. These are measures motivated by the Debreu (1951) coefficient of resource utilization,
which determine the maximum proportional uniform shrinkage in the endowments of the global
economy which is possible subject to the constraint that global utility (in the form of a global
social welfare function) is preserved as distortions are removed, and provides a measure of the
degree of inefficiency of the global economy. Using this type of globalization metric implies that,
in a two good, two factor model of the world economy, one computes the maximal proportional
shrinkage for each of the two factor endowments which preserves global utility as trade barriers
are removed.
As Riezman, Whalley and Zhang note for the single economy case, there is seemingly no
satisfactory way of choosing between these alternative measures since there is no single measure
which dominates all others. Each will yield a numerical measure of distance, and these metrics
will behave differently across alternative pairwise equilibrium comparisons. One measure may
indicate a large distance between equilibria and another small. The interpretation of such
measures thus becomes an issue. What can one conclude, for instance, if endowments change
by, say, 10% in a Debreu type measure in comparsions across equilibria while an excess demand
calculation implies say a 50% measure for the same tariff barrier. Our purpose is thus to
raise the issue of which distance measures to use in globalization debate and why, rather than
definitively resolve the issue of which measure to use in all circumstances.
To show more concretely how measures of distance can be constructed between globalized
and non-globalized equilibria, we take the simple case of a global economy with N countries,
2 produced goods and 2 factors of production in each. For this economy, there will be various
5
features which limit integration of national economies into the global economy, such as tariffs,
domestic taxes, quotas and other policy interventions. We assume these are present in the
observed trade restricted (tariff) equilibrium, but absent in the hypothetical globally integrated
equilibrium.
For country n, we assume two input CES production functions for the two goods which are
given by
Qnj = φnj

 σnj −1
σnj

f =L,K
δnjf Tnjf
 σnj
 σnj −1

,
j = 1, 2
(1)
where Qnj denotes output of the j-th industry, φnj is the scale or units parameter, δnjf is the
distribution parameter ( f =L,K δnjf = 1), Tnjf is the labor and capital factor input (f = L, K),
and σnj is the elasticity of factor substitution.
The factor demand functions derived from cost minimization for these production functions
(1) are:
Tnjf =
Qnj
φnj

 σnj
−σnj  1−σnj  1−σnj
Dnf Dnf
δnjf ,
 
δnjf
δnjf j = 1, 2
f = L, K
(2)
f =L,K
where Dnf is the price of factor f (per-unit factor costs for the industry).
World prices for goods are P0j and are exogenous to the model, and the tariff rate on imports
of goods j is rnj . rnj > 0 if good j is imported (Xnj > Qnj ), and rnj = 0 if good j is exported
(Xnj ≤ Qnj ). The domestic price of good j is Pnj = (1 + rnj )P0j .
On the demand side of the economy we consider a representative consumer with a CES
utility function given by
Un =



j=1,2
1
σn
σn −1
σn
αnj Xnj
 σn
 σn −1

(3)
where Xnj is the quantity of good j demanded by the consumer, αnj is the share parameter
( j=1,2 αnj = 1), and σn is the substitution elasticity in the consumer’s CES utility function.
Consumer income has three parts, endowment income f =L,K Dnf Wf , tariff revenue Rn =
j=1,2 rnj P0j Znj and an exogenous foreign resource transfer Bn =
j=1,2 P0j Znj reflecting the
financing of any trade imbalance 3 .
In =
Dnf Wnf + Rn + Bn
(4)
f =L,K
3
We incorporate the trade imbalance in this way since typically actual economy data used in model calibration
will not be consistent with zero trade balance.
6
where Wf is the consumer’s endowment of labor and capital, Znj = Xnj − Qnj is the import
and export of good j (excess demands for goods).
The consumer’s budget constraint in country n is
Pnj Xj = In
(5)
j=1,2
where Pnj is the consumer price for good j.
Demand functions from utility maximizing behaviour are
Xnj =
σn
Pnj
αnj In
1−σn ,
j =1,2 αnj Pnj j = 1, 2
(6)
The equilibrium conditions in this model are that world market demand equals world market
supply for goods and also for factors in each country, and profits are zero in each industry. These
equilibrium conditions are that
[1] Demands equal supply for goods in the world
Xnj =
n=1,··· ,N
Qnj ,
j = 1, 2
(7)
n=1,··· ,N
[2] Demands equal supply for factors in each country
Tnjf = Wnf ,
f = L, K,
n = 1, · · · , N
(8)
j=1,2
[3] Zero profit conditions hold in both industries in each country
Dnf Tnjf = Pnj Qnj ,
j = 1, 2,
n = 1, · · · , N
(9)
f =L,K
In this global economy model, an equilibrium is characterized by factor prices DnL and
DnK , and world prices P01 and P02 , such that equations (7) - (9) hold.
From equations (2) and (9) we have
Dnf 1−σnj
δnjf
= [φnj Pnj ]1−σnj ,
δnjf
j = 1, 2
(10)
f =L,K
Defining excess demand of goods as
Znj = Xnj − Qnj ,
j = 1, 2
(11)
and excess demands for factors as
Ynf =
Tnjf − Wnf ,
j
7
f = L, K
(12)
the equilibrium conditions can be rewritten as [1] excess demands for goods in the world are
zero, i.e.
n=1,··· ,N Znj = 0, j = 1, 2; [2] excess demands for factors in each country are zero,
i.e. Ynf = 0, f = L, K, n = 1, · · · , N ; and [3] zero profit conditions hold in both industries (i.e.
Equation (9)).
At such an equilibrium we can also show that the trade imbalance constraint is satisfied in
each country. Using the full employment condition (8) and zero profit conditions (9), we have
In =
Pnj Qnj + Rn + Bn
(13)
j=1,2
from equation (4). Rewriting the budget constraint (5) as
Pnj Xnj =
j=1,2
that is,
Pnj Qnj + Rn + Bn
(14)
j=1,2
(1 + rnj )P0j Znj =
j=1,2
Pnj Znj =
j=1,2
rnj P0j Znj + Bn
(15)
j=1,2
from the excess demand of goods equation (11), and
P0j Znj = Bn
(16)
j=1,2
which is the trade imbalance.
Where distance metrics of globalization for the world economy require a global social welfare
function (as for the Debreu type measures), we assume a set of share parameters for each
countries as follows
Q0n1 + Q0n2
0
0 .
n=1,··· ,N (Qn1 + Qn2 )
λn = (17)
Global utility is thus, for simplicity, assumed to be additive across the N countries using the
share parameters, i.e. U = n=1,··· ,N λn Un .
Into this economy, we can introduce a wide range of trade distorting instruments in each
country beyond tariffs, such as domestic taxes and quotas. All of these policy interventions will
limit the degree of integration of national economies into the wider global economy. Assuming for now that equilibria in this global economy are unique, this economy has two equilibria
for cases with and without country trade policy interventions. We label these restricted and
unrestricted economies as E (1) and E (2) . Typically we will have an observed equilibrium E (1)
in the presence of trade barriers and a model parametrization will be calibrated to this equilibrium data so as to be consistent with it. The other equilibrium E (2) will be computed as a
counterfactual equilibrium.
8
We can construct various distance measures between equilibria for different barriers applying
to imported goods in each country, including barrier elimination. These equilibria are charac(1)
(1)
(1)
(1)
(1)
(1)
(2)
(2)
(2)
(2)
(2)
(2)
terized by the variables (P0j , Dnf , Pnj , Qnj , Tnjf , Xnj ) and (P0j , Dnf , Pnj , Qnj , Tnjf , Xnj )
for n = 1, · · · , N , j = 1, 2 and f = L, K.
Riezman, Whalley and Zhang (2004) define simple normalized Euclidean distance measures between restricted and unrestricted small open economy equilibria for each country in
prices and quantities as
MDn =
MP n =
MQn =
MT n =
f =L,K
(1)
(2) 2
Dnf − Dnf
1
(m)
m=1,2
f =L,K Dnf
4
(1)
(2) 2
P
−
P
j=1,2
nj
nj
1
(m)
m=1,2
j=1,2 Pnj
4
(1)
(2) 2
Q
−
Q
j=1,2
nj
nj
1
(m)
m=1,2
j=1,2 Qnj
4
(1)
(2) 2
−
T
T
j=1,2
f =L,K
njf
njf
(18)
(19)
(20)
(21)
1
(m)
T
8 m=1,2 j=1,2 f =L,K njf
(1)
(2) 2
j=1,2 Xnj − Xnj
(22)
MXn =
2
1
(m)
X
4 m=1,2 j=1 nj
We compute global normalized Euclidean distance measures between restricted and
unrestricted global equilibria in prices and quantities as
(1)
(2) 2
−
P
P
j=1,2
0j
0j
MP 0 =
1
(m)
P
4 m=1,2 j=1,2 0j
2
(1)
(2)
j=1,2
n=1,··· ,N Qnj − Qnj
MQ =
2 1
(m)
m=1,2
n=1,··· ,N Qnj
j=1
4
2
(1)
(2)
X
−
X
j=1,2
n=1,··· ,N
nj
nj
MX =
2 1
(m)
m=1,2
n=1,··· ,N Xnj
j=1
4
9
(23)
(24)
(25)
We also construct global excess demand measures between the two economies: E (1) and
(1)
(1)
(1)
(1)
(1)
(1)
(2)
(2)
(2)
(2)
(2)
(2)
E (2) and their equilibria: (P0j , Dnf , Pnj , Qnj , Tnjf , Xnj ) and (P0j , Dnf , Pnj , Qnj , Tnjf , Xnj )
for n = 1, · · · , N , j = 1, 2 and f = L, K, are the associated variable values. We assume E (1) in
the presence of barriers is observed, and we then introduce the prices from E (1) into the model
parametrization supporting E (1) and compute excess demands. This procedure does not yield
an equilibrium solution to the model, but does suggest locally how large excess demands for
goods (trade) would be were trade barriers to be eliminated in all countries.
We construct excess demand measures in either factor space in each country, or in goods
(2)
space across all countries. In the factor space case, we introduce the production values Qnj
(1 )
into model supporting Economy E (1) , and evaluating excess demand functions for factors Ynf
(1 )
(2)
given by equation (12). Given Qnj , we solve for Tnjf from Equation (2), for j = 1, 2 and
f = L, K,
(1 )
Tnjf =
(2)
Qnj
φnj

−σnj
(1)
Dnf


δnjf

 
δnjf 
σnj
1−σnj  1−σ
(2)
 nj
Qnj (1)

= (1) Tnjf

δnjf Qnj
(1)
B  nf
f =L,K
(1 )
and generate the excess demands for factors Ynf
=
(1 )
j=1,2 Tnjf
(26)
− Wnf (f = L, K) as in
equation (12). This yields a factor excess demand measure between E (1) and E (2) for country
n as
(1)
RF n
=
(1)
f =L,K
(1 )
Dnf |Ynf |
(1)
(1 )
f =L,K Dnf Tnjf
j=1,2
(27)
In this evaluation of factor excess demands full employment conditions do not hold, and so
consumer’s income is given by
In(1 ) =
f =L,K
(1 )
(1 )
(1)
Dnf Wnf +
j=1,2
(1)
(1 )
rnj P0j Znj + Bn
(28)
(2)
where Znj = Xnj − Qnj .
(1 )
Solving for consumption Xnj from Equation (6), for j = 1, 2,
(1)
(1)
(1 )
(2)
αnj
D
W
+
r
P
[X
−
Q
]
+
B
0j
n
nf
f =L,K nf
j=1,2 nj
nj
nj
(1 )
Xnj =
(1) σn (1) 1−σn
[Pnj ]
j αnj [Pnj ]
(29)
we also generate a global excess demand distance measure between E (1) and E (2) using goods
(1 )
excess demands Znj
from Equation (11). This yields a goods excess demand measure of
distance for country n as
(1)
RGn
=
(1) (1 )
j=1,2 Pnj |Znj |
(1) (1 )
j=1,2 Pnj Xnj
10
(30)
Global excess demand globalization measures in goods space between economies are
(2)
calculated by introducing the world prices P0j into model supporting Economy E (1) , to get
(1 )
the domestic good prices Pnj ,
(1 )
(1)
(2)
Pnj = (1 + rnj )P0j
(31)
(1 )
(1 )
we then solve for factor prices Dnf from equation (10), compute production Qnj and factor
(1 )
(1 )
use Tnjf from equations (2) and (8), and obtain consumer income In
(1 )
(1 )
and consumption Xnj
in equations (31) and (32), then goods excess demands Znj from Equation (11). This yields a
goods excess demand measure of distance between E (1) and E (2) as
(1)
RG
j=1,2
= j=1,2
(1)
(1 )
(1)
(1 )
n=1,··· ,N
Pnj |Znj |
n=1,··· ,N
Pnj Xnj
.
We can finally construct Debreu type shrinkage measures of distance between the two
(1)
(1)
(1)
(1)
(1)
(1)
global equilibria: E (1) and E (2) and their characteritics: (P0j , Dnf , Pnj , Qnj , Tnjf , Xnj ) and
(2)
(2)
(2)
(2)
(2)
(2)
(P0j , Dnf , Pnj , Qnj , Tnjf , Xnj ) for n = 1, · · · , N , j = 1, 2 and f = L, K. To do this we use
(2)
P0j in the model specification supporting Economy E (1) , and compute a global equilibrium for
the case where tariffs are eliminated in all countries and there is a supporting endowment of
(1 )
(1)
(1)
factors Wnf = [1 + RDn ]Wnf which yields unchanged utility for the representative consumer
(1)
n. RDn yields the Debreu type shrinkage measure of distance for country n between the two
equilibria E (1) and E (2) . For the global economy, there is a scalar supporting endowment of
(1 )
(1)
(1)
(1)
factors Wnf = [1 + RD ]Wnf (for all n = 1, · · · , N ) which yields unchanged global utility. RD
yields the Debreu type shrinkage measure of distance between the two equilibria E (1) and E (2) .
11
3
Applying Globalization Distance Metrics to the Global Economy
To investigate how the globalization distance measures set out above behave in practice, and
what numerical measures of globalization distance are implied for the world economy, we have
used the simple 2 good, 2 factor global trade model set out above to construct globalization
distance measures between the actual world economy and hypothetical fully integrated world
with no barriers between countries involving a sample of OECD economies and a residual rest
of world (ROW). We treat Australia, Canada, Germany, Italy, Japan, Korea, Mexico, Norway,
UK, US as separate OECD economies which differ in size, their trade patterns, their levels of
development, and their degree of openness and add in a residual rest of the world yielding a
10 region model. We use the OECD STAN database for 2000 which provides consistent data
on consumption, production, and trade for all OECD economics. Into this we also introduce
measures of average tariff rates on imports taken from OECD sources. We find this OECD
data simpler to use and in some ways more applicable to our needs than the GTAP data base
currently widely used by trade equilibrium modellers.
We construct an initial base case global data set reflecting an equilibrium for each of these
economies in the presence of domestic trade restrictions, which for simplicity we limit to tariffs.
The equilibrium data for each economy is set out in Table 1. We have generated this by
taking information from the STAN database in value terms in domestic currency from which
we assemble consumption, production, factor by sector, and net trade for each country for the
year 2000.
12
13
73,354
55,593
33,058
K3 - M
94,772
62,681
104,127
20,227
198,899
82,908
112,580
11,880
310,640
15,870
423,220
27,750
Germany
49,096
M
122,450
M
197,570
48,865
-1,329
-34,043
303,129
76,083
-120,091
48,333
182,608
62,088
-37,996
26,554
98,397
27,220
122,207
8,314
220,604
35,534
Italy
92,936
17,376
-19,178
9,604
52,608
5,390
59,506
2,382
112,114
7,772
Japan
0.673166
0.105931
M
0.000000
0.000000
0.000000
0.208675
0.923392
105,882
68,392
-57,401
42,072
103,613
23,205
59,670
0.000000
0.158008
0.009275
3,115
163,283
0.000000
0.368725
0.000884
0.348223
0.000000
0.105719
1,159,129
141,163
145,532
-131,025
702,358
221,083
311,239
51,105
1,013,597
272,188
Mexico
0.032875
0.000000
0.113496
220,763
52,604
78,985
-305,955
44,118
333,498
97,660
25,061
141,778
358,559
Norway
0.041334
0.000000
1.515850
190,470
33,784
36,799
-0,325
43,188
27,946
110,483
6,163
153,671
34,109
UK
0.033641
0.109535
1.000000
1,841,611
356,465
321,348
89,103
538,482
178,172
981,781
89,190
1,520,263
267,362
US
0.000000
0.000000
1.000000
1,285,722
61,297
-34,893
-155,999
530,964
157,434
789,650
59,861
1,320.615
217,295
ROW
2 for the underlying data used to generate tariff averages reputed here.
4 See Table
3 In this table, M and NM denote Total Manufacturing and
Non Manufacturing (Agriculture, Hunting, Forestry and Fishing; Mining and Quarrying); L and K denote Labour and Capital.
JPY 109 , KRW 109 , MXP 106 , NWK 106 , GBP 106 , USD 106 , and USD 106 .
2 These value units in domestic currency are AUD 106 , CAD 106 , EUR 106 , EUR 106 ,
0.000000
0.208675
0.923392
1 Sources: OECD STAN database plus Table 2.
0.000000
NM
Initial Tariff Rate on Imports4
0.581567
Korea
26,320
Exchange Rate (How Many USDs Are Equal to 1 Unit Local Currency)
32,172
NM
Value of Consumption
-23,421
NM
Value of Net Trade (Imports - Exports)
45,088
40,296
10,505
K - NM
3
L -M
3
L3 - NM
Value of Factor Use
M
3
NM3
Value of Output
Canada
and a Residual ROW1 (Country Data in Domestic Currency2 )
Value Data Reflecting Assumed 2000 Benchmark Equilibria for a Sample of OECD Economies
Australia
Table 1
We use a two sector classification for each economy, and we first ignore all service related
and non-tradable transactions such as utilities, government activity, retailing, wholesaling, distribution, banking, and financial services. We consider two aggregate traded goods sectors
which we take to reflect manufacturing and non-manufacturing activity. From the STAN data,
“total manufacturing” is taken as manufacturing and “agriculture, fishing, forestry, and mining
/ quarrying” are taken as non-manufacturing. The 2 × 2 case requires as many factors as goods
in the model, and so this is a relatively easy case to implement. Subsequent applications of this
approaching can easily (but more tediously) be applied to multisector multifactor models and
the change of level aggregation in data will clearly affect the numenical value of globalization
distance measures.
STAN data give value added according to this sectoral classification, and also provide data
on compensation of employees by sectors. The return to capital by sector in each country is
constructed by residual as the difference between the two. We make the strong assumption
that the output of each sector is given only by the value added originating in the sector, and
we ignore all intermediate transactions. This yields data on output and factor use by sector
in value terms for each country for our benchmark year. This data is in value terms, and to
produce equilibrium data on both prices and quantities we need to adopt a units convention for
the measurement of both goods and factors. We follow the convention attributed to Harberger
(1962) and discussed in Shoven and Whalley (1992) of assuming unitary prices for factors, and
unitary world prices for goods in the trade distorted equilibrium. This yields domestic prices
for imports as one plus the tariff rate.
We use the trade data in STANs on a net trade basis and net out imports and exports
(again in value terms) by good (for our 2 good classification) for each country. This yields
consumption as production plus net trade. This substantially reduces trade volumes as they
appear in each country model relative to published trade data. Most of the OECD economies
we consider are net exporters of manufactured goods. Trade balance does not hold for this net
trade data by country since some countries have trade surpluses and other (notably the US)
have trade deficits. One procedure is to modify the data to force trade balance; another is to
use a model which incorporates a fixed trade imbalance (which is non-zero). We choose the
latter procedure. For the US, this yields the feature that both goods are imported, and financed
by foreign resource transfers supporting the observed trade imbalance.
We use tariff rate data from OECD sources as our trade barrier representation in the re-
14
stricted equilibrium. This data (also used by Riezman, Whalley and Zhang (2004)) is presented
in Table 2 which also reports the tariff data we have used to produce on for our barrier estimates. This data is from OECD Sources on bound tariff rates by Harmonized Nomenclature
section headings, and gives the fraction of line items in specified tariff lines falling in numerical
ranges of tariff rates. There is a considerable literature on constructing tariff averages, which
for simplicity we ignore. We use statutory rather than effective tariff rates, and we have aggregated this data using simple means for in sample ranges. We do not employ trade weighted
average, nor use applied rather than bound tariff rates.
The benchmark equilibrium in Table 1 is taken as observed for our representation of the
global economy. We calibrate our global model to this data which we assume for now to
be generated in the presence of tariffs as the only trade barrier which creates distance from
full integration. We then apply the procedures set out in the previous section and construct
globalization metrics for the world economy.
The procedures set out in Section 2 are relatively simple to implement, but there are a
number of issues of detail which arise. One is that in a model in which factors are fully mobile
within each economy production frontiers in each economy will be close to linear for conventional
fractional forms such as Cobb Douglas and CES unless share and substitution parameters differ
sharply across sectors. This means that even for small barrier changes specialization will occur
and so globalization distance measures may only be able to be computed for this model for
modest barrier reductions rather than barrier eliminations.
In reality there are many barriers which limit integration into the global economy including
other trade measures (quotas, dumping and countervailing daties), national standards, differential regulation of financial institutions, transportation regulation, agricultural policies, and
many others. Each of these would need an explicit model representation which world sharply
differ from a representation by an advalorem equivalent tariff if they were to be sensibly incorporated into such analyses. Extensions of this approach can be used to analyze now these
barriers also affect globalization measures. We have not done so since our purpose is to illustrate a general approach to constructing globalization metrics rather than provide definitive
estimates.
15
16
16.2
1.6
0.0
44.0
4.2
1.4
18.7
19.6
28.1
17.7
2.5
11.6
1.7
0 − 5%
5 − 10%
10 − 15%
15 − 20%
20 − 50%
> 50%
Duty Free
0 − 5%
5 − 10%
10 − 15%
15 − 20%
20 − 50%
> 50%
0.0
0.2
7.1
8.7
39.5
16.4
28.6
0.9
6.3
26.6
0.0
0.3
23.9
42.1
Canada
0.0
0.2
0.4
7.0
27.3
43.0
22.2
17.0
11.2
12.2
7.7
9.3
16.1
26.5
Germany
10.5931
M
5.7168
8.1469
4.1334
20.8675
9.6
10.3
22.0
1.2
6.7
19.1
31.0
0.0
0.2
0.4
7.0
27.3
43.0
22.2
4.1334
20.8675
2.3100
15.8008
0.0
0.4
0.2
2.7
14.5
28.4
53.8
Manufacturing
17.0
11.2
12.2
7.7
9.3
16.1
26.5
11.8107
36.8725
0.2
10.5
8.2
36.8
23.8
7.1
13.5
38.6
8.1
7.2
19.1
16.6
8.4
2.2
Korea
0.0
99.2
0.3
0.0
0.2
0.0
0.2
84.8
0.3
0.0
8.4
3.0
3.3
0.1
Mexico
34.8223
65.1276
2 Sources: Tariffs and Trade: OECD query and reporting system, OECD 2000.
1 Note: Calculations are report 6 - digit HS section headings.
10.7450
NM
Japan
Non Manufacturing
Italy
Calculated Country Average Tariff Rates (Percentage)
32.6
Australia
Duty Free
Ranges
Binding
Tariff
by Country Used to Calculate Country Tariff Rates in Table 11
2
3.2875
12.8729
0.2
0.0
0.0
10.6
13.7
31.4
44.1
4.2
0.3
14.3
56.4
0.4
0.9
23.4
Norway
4.1334
20.8675
0.0
0.2
0.4
7.0
27.3
43.0
22.2
17.0
11.2
12.2
7.7
9.3
16.1
26.5
UK
US
3.3641
10.9535
0.0
0.8
1.3
5.6
15.6
39.6
37.2
0.4
4.7
48.1
0.1
2.0
16.7
27.9
OECD Data on Tariff Intervals by HS Section (Post-Uruguay Round Bound Rates)
(% of Tariff Nomenclature Section Headings in Specified Rate Ranges by Country)
Table 2
Table 3 presents the calibrated model parameter values for our global model, along with
model data on endowments and tariff rates. The calibration procedures we use are set out
in Dawkins, Srinivasan and Whalley (2001). To implement calibration we rely on a simple
literature search to generate substitution elasticities by sector by country, and we use values
roughly consistent with those reported in Piggott and Whalley (1985) and Hammermesh (1993)
of 2.0 in non manufacturing and 0.5 in manufacturing. We use each country model to compute
a new global equilibrium with reduced country tariffs. We are able to then construct sum
of squares distance measures between these equilibria, as set out in previous section. We are
also able to use the model parameterizations supporting each equilibrium to construct excess
demand distance measures, and Debreu type shrinkage measures of distance of economies from
full integration (globalization).
17
18
1.804019
M
1.998893
1.858949
0.524728
0.674450
0.475272
L-M
K - NM
K-M
0.488235
0.637729
0.511765
0.362271
0.796049
M
78,146
K
157,453
124,354
0.801712
0.198288
124,460
326,510
0.791659
0.208341
0.375783
0.463866
0.624217
0.536134
1.883737
1.646108
Germany
125,617
130,521
0.737189
0.262811
0.472938
0.644054
0.527062
0.355946
1.994158
1.527880
Italy
10.5931
M
0.0000
0.0000
0.0000
20.8675
0.0000
20.8675
-35,372.0
-80,102.6
-16,026.5
-10,884.5
0.0000
-26,662.9
0.0000
36.8725
1.25
-23,081.5
34.8223
0.0000
923,441
362,344
0.898458
0.101542
0.600354
0.675316
0.399646
0.324684
1.425990
1.781035
Mexico
-229,483.9
3.2875
0.0000
377,616
122,721
0.808824
0.191176
0.401958
0.784851
0.598042
0.215149
1.864649
1.509935
Norway
35,013.3
4.1334
0.0000
71,134
116,646
0.850640
0.149360
0.384699
0.680453
0.615301
0.319547
1.823638
1.769515
UK
352,576.6
3.3641
10.9535
716,654
1070,971
0.835407
0.164593
0.425482
0.585645
0.574518
0.414355
1.892864
1.751178
US
-190,891.4
0.0000
0.0000
688,398
849,511
0.954495
0.045505
0.450550
0.618571
0.549450
0.381429
1.980627
1.893515
ROW
from abroad.
2 Footnote from Table 1.
1 Model tariff rates on exports are set equal to zero, the US imports both goods, with the trade imbalance financed by a resource transfer
20,972.4
Foreign Resource Transfers in Domestic Currency2
0.0000
NM
Initial Tariff Rate on Imports1 (Percentage)
Substitution Elasticities in Consumption (All Countries)
2.0
M
126,818
62,785
0.588701
0.411299
0.568545
0.731858
0.431455
0.268142
1.963106
1.202613
Korea
2.0
15.8008
57,998
61,888
0.837555
0.162445
0.484604
0.600681
0.515396
0.399319
1.998105
1.659805
Japan
NM
Substitution Elasticities in Production (All Countries)
50,801
L
Initial Endowment
0.203951
NM
Share Parameters in Preferences
0.325550
L - NM
Share Parameters in Production
1.782958
NM
Scale Parameters in Production
Canada
Calibrated and Other Model Parameter by Country Reproducing the Equilibrium Data in Table 1
Australia
Table 3
4
Some Calculations of Global Distance Measures
We have made calculations of globalization distance measures for the world economy for the
year 2000 using ten OECD countries (Australia, Canada, Germany, Italy, Japan, Korea, Mexico,
Norway, the UK, and the US) and a residual rest of the world. To make these calculations we
first calibrate our global model as described above and then proceed to compute a barrier
change equilibrium as our counterfactual in each case. We then compare initial and barrier
change equilibria and construct the globalization measures for the world economy set out in
Section 2.
As we note above, due to the near linearity of the production frontier in the 2 × 2 case
(see Johnson (1966), and Abrego and Whalley (2001)) specialization occurs as the equilibrium
outcome on most countries in the tariff elimination case, The US, The UK, and Australia are the
exceptions. We therefore compute globalization metrics for a common percentage reduction in
tariffs in all countries. We use 1.000%, 2.000%, 3.000%, 3.600%, 3.690%, and 3.698% common
reduction in tariff rates in all countries with 3.69803045848% as the maximal common reduction
computationally possible.
Table 4 presents the results. Just as in Riezman, Whalley and Zhang (2004) there are many
globalization measures (and equilibrium metrics in general) available for the global economy
and their numerical behavior differs. Euclidean distance measures which differing one from
another are approximately linear in the depth of tariff cut. Excess demand measures, however,
are smaller in some case for larger tariff costs.
Which measure to use, and how to interpret such measures when calculated, especially when
reported in popular press, is not clear. Not only do they vary in size, but they also vary in their
precentage changes across tariff reductions of different depth. Measures will also vary further
with the degree of disaggregation in models, the structural form of models, and the treatment
of factor flows and barriers. Hence, making sense of the numbers which currently circulate in
popular media as measures of globalization in popular media thus seems to be a hazardous
occupation, if one is to say how far equilibria are apart. This also presumes that equilibrium
rather than disequilibrium adequately characterizes the state which the global economy finds
itself in.
19
Table 4
Globalization Metrics for the World Economy for Common
1.000%, 2.000%, 3.000%, 3.600%, 3.690%, and 3.698%
Reductions in Tariff Rates in Each Country
1.000%
2.000%
3.000%
3.600%
3.690%
3.698%
Euclidean Distance Measures
MP0
0.000602
0.001219
0.001863
0.002235
0.002293
0.002298
MQ
0.007433
0.014704
0.021913
0.026035
0.026663
0.026718
MX
0.001043
0.002048
0.003026
0.003582
0.003666
0.003673
0.217444
0.215840
0.215598
0.215576
0.000820
0.000988
0.001013
0.001015
Excess Demand Measures
1
RG
0.222684
0.220066
2
RG
0.238726
0.251927
Debreu Shrinkage Measures
RS1
0.000269
0.000543
RS2
-0.000271
-0.000548
20
Table 5 compares country measures of globalization distance for both 1.000%, 2.000%,
3.000%, and 3.600% common tariff reductions generated by the global model set out in earlier
sections to separate country calculations of the type reported by Riezman, Whalley and Zhang
(2004) for the small open economy case. These measures differ sharply, highlighting the pitfalls in making separate country calculations of globalization distance rather than using global
measures. Table 5 also reveals competing influences on results. Because there is a common %
reduction in tariffs, countries with higher initial tariff (Mexico, for instance) have larger globalization metrics. But also in globally based measures metrics are smaller for larger economies
(the US, for instance) as smaller economies integrate into their markets and there is smaller
domestic price change as barriers are lowered.
21
22
Canada
0.001979
0.001981
Models
0.000000
0.000000
0.003988
0.003995
Models
0.000000
0.000000
0.006072
0.006088
Models
0.000000
0.000000
0.007331
0.007354
Models
0.000000
0.000000
Global Models
RF1
0.003572
2
RF
0.003578
Individual Economy
RF1
0.002198
2
RF
0.002200
Global Models
RF1
0.007154
RF2
0.007180
Individual Economy
RF1
0.004391
2
RF
0.004401
Global Models
RF1
0.010826
2
RF
0.010867
Individual Economy
RF1
0.006580
RF2
0.006602
Global Models
RF1
0.012951
2
RF
0.013034
Individual Economy
RF1
0.007891
2
RF
0.007922
Italy
Japan
Korea
Mexico
Norway
0.005149
0.005137
0.008417
0.008452
0.006023
0.006011
0.003370
0.003366
0.009464
0.009374
0.007339
0.007291
0.016998
0.017287
0.020927
0.021460
0.000221
0.000221
0.000637
0.000637
0.010340
0.010291
0.006700
0.006680
0.012088
0.012037
0.006707
0.006692
0.019181
0.018816
0.014808
0.014591
0.033514
0.034614
0.041104
0.042820
0.000441
0.000442
0.001281
0.001282
0.015572
0.015460
0.025079
0.025378
0.018195
0.018078
0.009957
0.009922
0.029163
0.028327
0.022332
0.021839
0.049211
0.051678
0.060125
0.063830
0.000662
0.000662
0.001941
0.001943
0.011946
0.011937
0.018732
0.018570
0.019377
0.019556
0.030037
0.030462
0.021880
0.021710
0.011901
0.011852
0.035285
0.034067
0.026922
0.026208
0.058456
0.061955
0.071339
0.076584
0.000795
0.000795
0.002338
0.002341
Common 3.600% Reduction in Tariff Rate for all Countries
0.009995
0.009948
0.016224
0.016350
Common 3.000% Reduction in Tariff Rate for all Countries
0.016776
0.016910
0.010883
0.010939
Common 2.000% Reduction in Tariff Rate for all Countries
0.003355
0.003350
0.005500
0.005515
Common 1.000% Reduction in Tariff Rate for all Countries
Germany
0.002040
0.002042
0.005217
0.005229
0.001700
0.001702
0.004332
0.004340
0.001134
0.001134
0.002862
0.002865
0.000567
0.000567
0.001424
0.001426
UK
(Common 1.000% and 2.000% Reductions in Tariff Rate for all Countries)
0.006980
0.006972
0.000451
0.000451
0.005813
0.005807
0.000410
0.000410
0.003871
0.003869
0.000329
0.000329
0.001933
0.001933
0.000180
0.000180
US
ROW
0.000000
0.000000
0.006483
0.006497
0.000000
0.000000
0.005369
0.005378
0.000000
0.000000
0.003524
0.003528
0.000000
0.000000
0.001747
0.001748
Comparing Globally Constructed and Individual Economy Based Measures of Globalization
Australia
Metrics
Table 5
Table 6 presents sensitivity analyses of computation of globalization metrics in Table 4. Here
we vary substitution elasticity in production between 0.75 and 4.00. The maximum percentage
reduction in tariff that are computationally feasible falls and so results across the case are not
directly comparable. Euclidean distance measures do not change as these are data not model
based. Excess demand measures seem relatively in sensitive, while Debreu shrinkage measures
vary more.
23
Table 6
Sensitivity Analyses of Global Metrics of Globalization
(Change Substitution Elasticities and Reproduce Table 4)
σnj = 0.750 for n = 1, · · · , N and j = 1, 2
1.000%
2.000%
3.000%
4.000%
5.000%
6.000%
Euclidean Distance Measures
M P0
0.000601
0.001207
0.001815
0.002428
0.003044
0.003664
MQ
0.002813
0.005616
0.008413
0.011201
0.013982
0.026756
MX
0.000403
0.000802
0.001197
0.001588
0.001975
0.002358
Excess Demand Measures
1
RG
0.224370
0.223330
0.222283
0.221311
0.220344
0.219371
2
RG
0.230539
0.235650
0.240762
0.245866
0.250961
0.256049
Debreu Shrinkage Measures
1
RS
0.000109
0.000214
0.000321
0.000427
0.000533
0.000638
2
RS
-0.000107
-0.000215
-0.000322
-0.000430
-0.000537
-0.000645
σnj = 1.500 for n = 1, · · · , N and j = 1, 2
1.000%
2.000%
3.000%
4.000%
4.900%
4.910%
Euclidean Distance Measures
M P0
0.000601
0.001212
0.001833
0.002464
0.003043
0.003048
MQ
0.005596
0.011110
0.016548
0.021913
0.026689
0.026735
MX
0.000787
0.001553
0.002299
0.003026
0.003666
0.003672
0.217435
0.215630
0.215610
0.000826
0.001015
0.001017
Excess Demand Measures
1
RG
0.223363
0.221344
0.219407
2
RG
0.235456
0.245445
0.255375
Debreu Shrinkage Measures
1
RS
0.000204
0.000410
0.000617
2
RS
-0.000205
-0.000413
-0.000623
24
Table 6
Sensitivity Analyses of Global Metrics of Globalization
(Change Substitution Elasticities and Reproduce Table 4) —- CONTINUED
σnj = 2.000 for n = 1, · · · , N and j = 1, 2
1.000%
2.000%
3.000%
3.600%
3.690%
3.698%
Euclidean Distance Measures
MP0
0.000602
0.001219
0.001863
0.002235
0.002293
0.002298
MQ
0.007433
0.014704
0.021913
0.026035
0.026663
0.026718
MX
0.001043
0.002048
0.003026
0.003582
0.003666
0.003673
0.217444
0.215840
0.215598
0.215576
0.000820
0.000988
0.001013
0.001015
Excess Demand Measures
1
RG
0.222684
0.220066
2
RG
0.238726
0.251927
Debreu Shrinkage Measures
1
RS
0.000269
0.000543
2
RS
-0.000271
-0.000548
σnj = 3.000 for n = 1, · · · , N and j = 1, 2
1.000%
2.000%
2.400%
2.470%
2.474%
Euclidean Distance Measures
MP0
0.000605
0.001234
0.001491
0.001536
0.001539
MQ
0.011067
0.021745
0.025920
0.026645
0.026687
MX
0.001549
0.003011
0.003575
0.003672
0.003678
0.215849
0.215566
0.215550
0.000982
0.001011
0.001013
1.850%
1.858%
1.8589%
Excess Demand Measures
1
RG
0.221361
0.217454
2
RG
0.245237
0.227728
Debreu Shrinkage Measures
1
RS
0.000401
0.000814
2
RS
-0.000404
-0.000740
σnj = 4.000 for n = 1, · · · , N and j = 1, 2
1.000%
1.800%
Euclidean Distance Measures
MP0
0.000609
0.001119
0.001151
0.001156
0.001156
MQ
0.014648
0.025863
0.026552
0.026662
0.026649
MX
0.002045
0.003574
0.003666
0.003681
0.003682
0.215584
0.215541
0.215536
0.001007
0.001011
0.001012
Excess Demand Measures
1
RG
0.220080
2
RG
0.251711
0.215854
Debreu Shrinkage Measures
1
RS
0.000535
2
RS
-0.000540
0.000979
25
5
Concluding Remarks
In this paper we discuss metrics of globalization for the world economy constructed as
distance measures between barrier restricted and globally integrated trade equilibriua. This
contrasts with an earlier paper of ours reporting globalization metrics for single economies
treated as price takes in world markets. We emphasize that literature on equilibrium metrics
in general is limited, and so our paper is also a contribution to part of a wider discussion of
metrics of distance across equilibria in other circumstances.
We report measures generated using a global model involving 10 OECD countries and
residual rest of world using 2000 data applied to simple 2 × 2 country models. These suggest
both substantial differences in measures for the global economy and differences across countries
in country measures for the same barrier reduction. As such our results may seem on the whole
negative, but they again illustrate the pitfalls involved with measures currently reported in
populist magazines such as the Economist and Foreign Policy.
Importantly, we show how metrics produced from an explicit global model, and those from
single economy models / where each economy is assumed to be a price taker) will differ. Large
economies integrating with small economies experience little relative price change and hence are
already integrated into the global economy (their own). Small economies degree of integration
is influenced as much if not move by foreign barriers than their own. Seemingly an integrated
model of the global economy is needed to sensibly discuss globalization, and metrics only
meaningfully exist for the world, not individual economics.
26
References
[1] Arrow, K.J. and F. Hahn. (1971) General Competitive Equilibrium. Mathematical
Economics Texts 6, San Francisco: Holden-Day, Inc.
[2] Cable, V. (1995) The Diminished Nation-State: A Study in the Loss of Economic
Power.
Daedalus (Spring 1995).
[3] Cable, V. and A. Bressand. (2000) Globalization: Rules and Standards for the
World Economy.
Chatham House Papers, Thomson Learning.
[4] Debreu, G. (1951) The Coefficient of Resource Utilization.
Econometrica 19 (3), 273 -
292.
[5] Dawkins, C., T.N. Srinivasan, and J. Whalley. (2001) Calibration.
in J.Heckman and
E. Leamer (Eds) Handbook of Econometrics, Volume 5, 3653 - 3703. New York, N.Y.:
Elsevier Science Publishing. Co.
[6] Foreign Policy, Measuring Global.
March - April 2004, Page 54 - 69.
[7] Hamilton, R.W. and J. Whalley. (1984) Efficiency and Distributional Implications of
Global Restrictions on Labour Mobility: Calculations and Policy Implications. Journal of
Development Economics 14 (1-2), 61-75.
[8] Hammermesh, D. (1993) Labour Demand Elasticities. The World Economy 24 (1), 15 30.
[9] Harberger, A.C. (1962) The Incidence of the Corporate Tax. Journal of Political Economy
70 (3), 215 - 240.
[10] Irwin, D.A. (2001) Tariffs and Growth in Late Nineteenth Century America. The World
Economy 24 (1), 15 - 30.
[11] Kehoe, T. (1991) Computation and Multiplicity of Equilibria.
in W. Hildenbrand and
H. Sonnenschein (Eds) Handbook of Mathematical Economics, volume 4, 2049-2143.
North Holland.
[12] Mas-Colell, A. (1985) The Theory of General Economic Equilibrium - A Differentiable Approach.
Econometric Society Monographs 9, Cambridge University
Press.
27
[13] Piggott, J.R. and J. Whalley. (1985) UK Tax Policy and Applied General Equilibrium
Analysis.
Cambridge: Cambridge University Press.
[14] Riezman, R., J. Whalley and S. Zhang. (2004) Metrics Capturing The Degree to Which
Individual Economies Are Globalized. Forthcoming in Dissecting Globalization (Edited
by John Whalley), MIT Press for CES-ifo, Venice, 2004.
[15] Shoven, J.B. and J. Whalley. (1992) Applying General Equilibrium.
Cambridge
University Press.
[16] Whalley, J. (2004) Globalization and Values.
Paper Prepared for CES-ifo Conference
“Dissecting Globalization”, Vince, July 21 - 23, 2004.
[17] Whalley, J. and S. Zhang. (2004) Surfing the Manifolds of An Economy with Five Equilibria.
Research Paper, Department of Economics, The University of Western Ontario.
28