Estimating Aggregate Capital Stocks.pdf

8/10/2019 Estimating Aggregate Capital Stocks.pdf

1/42

Diskussionspapierreihe

Working Paper Series

Department of Economics

Fchergruppe Volkswirtschaftslehre

ESTIMATINGAGGREGATECAPITAL

STOCKSUSINGTHEPERPETUALINVENTORYMETHOD

NEWEMPIRICALEVIDENCE

FOR103 COUNTRIES

MICHAELBERLEMANNAND

JAN-ERIKWESSELHFT

Nr./ No. 125

OCTOBER2012


2/42

Autoren / Authors

Michael Berlemann

Helmut Schmidt UniversityDepartment of Economics

Holstenhofweg 85, 22043 Hamburg, [email protected]

Jan-Erik Wesselhft

Helmut Schmidt UniversityDepartment of EconomicsHolstenhofweg 85, 22043 Hamburg, [email protected]

Redaktion / Editors

Helmut Schmidt Universitt Hamburg / Helmut Schmidt University HamburgFchergruppe Volkswirtschaftslehre / Department of Economics

Eine elektronische Version des Diskussionspapiers ist auf folgender Internetseite zu finden/An electronic version of the paper may be downloaded from the homepage:http://fgvwl.hsu-hh.de/wp-vwl

Koordinator / Coordinator

Julia [email protected]


3/42

Helmut Schmidt Universitt Hamburg / Helmut Schmidt University HamburgFchergruppe Volkswirtschaftslehre / Department of Economics

Diskussionspapier Nr. 125Working Paper No. 125

Estimating Aggregate Capital Stocks Using the

Perpetual Inventory Method

New Empirical Evidence for 103 Countries

MICHAELBERLEMANNJAN-ERIKWESSELHFT

Zusammenfassung/ Abstract

The lack of internationally comparable capital stock data has been a major obstacle toempirical studies of the contribution of the capital stock to economic growth. In this paper, weprovide estimations of aggregate capital stocks for 103 countries in 2010. Depending on dataavailability the time series of the sample countries start in between 1960 and 1991. The

estimation is based on World Bank investment data and applies a unified approach ofapplying the Perpetual Inventory Method. The data can easily be extended for more recentyears as soon as new data is available.

JEL-Klassifikation / JEL-Classification: O47

Schlagworte / Keywords: aggregate capital stock, investments, perpetual inventory method


4/422

1 Introduction

In theoretical models of economic growth the physical capital stock, consisting of e.g.

machinery, buildings and computers, is one of the major input factors of the production

function. In order to study the contribution of the existing capital stock to aggregate output,

data on the capital stock is necessary. However, since the capital stock of a country is not

easily observable, data on the development of the capital stock has been unavailable for

most countries for a considerable time.

Nowadays, at least many industrial countries spend substantial effort on measuring

their capital stocks.3However, although international standards of measuring capital stocks

slightly evolve, the applied methods differ from case to case uite substantially.4 !s a

conseuence internationally comparable datasets are yet widely unavailable. "hile the

#$%& maintains a database of international capital stock data for its member countries, the

data is a mixture of data collected from the national statistical offices and own estimations

of the #$%&. 'he #$%& therefore recommends careful usage of the data for international

comparisons.5

'he lack of internationally comparable capital stock data has been a major obstacle

to empirical studies of the contribution of the capital stock to economic growth. In the

absence of reliable capital stock data the scientific literature has often employed different

proxies for capital accumulation.6!s a prominent example (!))# *++-, and much of the

related literature thereafter, employed gross investment rates as a proxy for physical capital

accumulation. "hile in the absence of reliable measures of the capital stock the use of

these proxies is an acceptable alternative, the construction of capital stock data is surely the

superior method. However, due to the fact that constructing capital stock data is a time

consuming task, most of the related literature has yet relied on the proxy approach.

!gainst the background of the considerable efforts to construct capital stock data it is

not too surprising that only a few attempts have yet been made in the literature to generate

/ ! documentation of the system of capital stock measurement in the 0nited 1tates is reviewed in

(0)$!0 # $%#N#3I% !N!451I1*266/-, the %anadian method in 1'!'I1'I%1 %!N!&!*266+-. or a description

of the methods of measuring the 7erman capital stock, see 1%H3!4"!11$)and 1%HI&4#"18I*2669-.:1%H)$5$) $' !4. *26++-, p. 2.

;1%H)$5$) $' !4. *26++-.

9($NH!(I(and 1


5/42


6/42:

employed methodology comes at the advantage that the dataset can easily be extended to

more recent years as the data becomes available.

'he paper is constructed as follows. 1ection 2 introduces the


7/42;

K=1 K+ I)epeatedly substituting this euation for the capital stock at the beginning of period

t-1,

K, leads toA

K= 1 I 'hus, the capital stock in period t is a weighted sum of the history of capital stock

investments. 'he weights result from the geometric depreciation function.

#bviously, calculating the actual capital stock in an accurate manner reuires to have

a complete time series of past investments. or many countries time series of investment

data are available for at least a certain number of years. However, these time series typically

cover only the *very- recent part of the capital stock history. 7iven the available time series

of investments is incomplete, we nevertheless can calculate the current capital stock Kaccurately whenever the initial capital stock at the beginning of the investment time series,, is knownA

K=1 K+ 1 I 'hus, in order to be able to apply the


8/429

3 Implementations of the Perpetual Inventory Method

#ver the years, various researchers have used the


9/42>

euilibrium, a shortterm investment shock in the first period of the available timeseries of

investments would lead to a strongly biased initial capital stock estimate.

!ware of this problem,H!)($)7$)*+>=- uses threeyear averages instead of a single

year to generate more stable and reliable capital stock estimates. In a later application of the

1teady 1tate !pproach, N$H)0and &H!)$1H"!)*+/- proposed an alternative procedure. In

order to generate a reliable initial value of the investment time series they regress the time

series of log investments on time and then use the fitted value for the first period to

calculate the initial capital stock.

3.2

Disequilibrium Approach

! second approach of estimating the initial capital stock goes back to 7)I4I%H$1*+=6-

and was used and further refined by $N$%Hand &$ 4! 0$N'$*2666-. 1imilar as the 1teady

1tate !pproach, the reasoning of this method bases on the neoclassical growth model. !s

outlined earlier, the capital stock can be written as

+=+=

K

t

GDP

t

t g

I

g

I

K 1 .

$N$%Hand &$ 4! 0$N'$*2666- argue that the growth rate of the capital stock can

be approximated by the growth rate of investments, i.e.

+

I

tt

g

IK

1.

However, different from the approaches in the tradition of H!)($)7$) *+>=-,$N$%Hand &$ 4! 0$N'$ *2666- argue that an economy typically is outside its longterm

euilibrium. rom their point of view it is more reasonable to assume that economies are

most of the time on their adjustment path towards euilibrium. 'hroughout this adjustment

process investment and capital accumulation tend to follow a systematic pattern. $N$%H

and &$ 4! 0$N'$ *2666- therefore propose to use data for longer timeperiods to estimate

the initial capital stock. 3ore precisely they use a Hodrick


10/42=

endpoints they drop the first ; annual observations of the smoothed investment timeseries.

!s proxy for the growth rate of investments they then use the average of the first ten

observations.

3.3 Synthetic ime Series Approach

! third procedure of estimating the initial capital stock goes back to !%#(, 1H!)3!

and 7)!(#"185 *+>- and was further refined by 8!3


11/42

residential assets and 2.; percent for government assets. or the subseuent years +9+ to

266+ he assumes the rate for private nonresidential assets to increase gradually from :.2;

percent to =.; percent, for government assets from 2.; to :.6 percent, thereby applying the

formula

= !"!#

$%

'he depreciation rate for private residential assets is held constant at +.; percent.

igure + shows the earlier described capital depreciation scheme applied by 8!3


12/42+6

turns out to be negative in the period*s- which are used to calculate the initial capital stock.

"henever the absolute value of the growth rate is considerably larger than the rate of

depreciation, the term g7&


13/42++

investments It+, the longterm growth rate of Investments gI and the rate of capital

depreciation LA

+

I

t

to

g

IK 1

However, we deviate from the procedure of $N$%H and &$ 4! 0$N'$ *2666- in

three respects.

irst, we do not use a filter to estimate the initial investment value. In order not to

lose any investment information we instead follow the idea of N$H)0and &H!)$1H"!)*+/-

to calculate the initial investment value It+from a regression approach. "e therefore use the

whole time series of investments, ranging from time t2to '. In order to do so, we regress the

time series of log investments ln*Ii,t- for any country i on time t. 'hus, we estimate the

euation

tiiiti tI ,,ln ++=

using the #41 method. In a next step we calculate the fitted value for period t +,

thereby using the estimated parameters Miand i, i.e.

,8 :; = ?.!fter transforming the fitted value using the exponential function we end up with a

time series of investments ranging from t+to '. "e then use the first *and thus the fitted-

value of this time series to calculate the initial capital stock in period t6.

1econd, we deviate from $N$%H and&$ 4! 0$N'$*2666- in the way of calculating

the growth rate of investments gI. Instead of using the mean of the investment time series

*or subsamples of the series- we employ the estimated parameter of ifrom the regression

as measure of trend investment growth.

'hird, we do not use a constant rate of depreciation in our approach, neither for the

calculation of the initial capital stock nor for the further construction of the time series of

capital stocks using the


14/42+2

1ince the capital depreciation schemes proposed by 8!3


15/42+/

from :.;J to =.;J in between +96 and 266+. "hile 8!3


16/42+:

database.13

'he resulting depreciation rate, which is shown in figure /, is then applied to all

sample countries.

Figure /: Assu(ed Aggregate Depreciation Rate of Gross Fixed Assets+ 190-2010

$ Sample "ountries and Data

#ur aim is to construct time series for capital stock data for a large sample of

countries. Instead of using #$%& data, which allow to differentiate between three classes of

capital investment but are only available for 22 #$%& countries, we rely on the aggregate

investment data provided by the "#)4& (!N8 in the "orld &evelopment Indicators *"&I-

database. "e extracted the gross fixed capital formation data with code N$.7&I.'#'.8& on

6/B26B26+2 from the database. 'he data includes land improvements *fences, ditches,

drains, and so on-Q plant, machinery, and euipment purchasesQ the construction of roads,

railways, and the like, including schools, offices, hospitals, private residential dwellings and

+/ 1ince our time series of depreciation rate has to date back to earlier years than +>6 and thus to

years for which nor disaggregate data are available, we decided to use the data of +>6 for these

years. or all years after +>6 the actual weighting factors are used.

6,66

6,;6

+,66

+,;6

2,66

2,;6

/,66

/,;6

:,66

:,;6

;,66

+;6 +;; +96 +9; +>6 +>; +=6 +=; +6 +; 2666 266; 26+6

!nnualcap

italdepreciationrateinJ

5ear


17/42+;

commercial and industrial buildings. !ccording to the +/ 1N!, net acuisitions of valuables

are also considered as capital formation. &ata are in constant 2666 0.1. dollars.+:

"hile the "&I database of the "#)4& (!N8contains aggregate investment data on a

large number of countries, the starting dates of the data differ heavily from country to

country. igure : illustrates aggregate data availability. or /6 countries, the investment

time series start out as early as in +96. 3ajor increases in the number of countries, for

which data is available are +9; *= countries-, +>6 *+9 countries-, +=6 *> countries- and

+6 *+: countries-. 'he +: countries added in +6 are primarily $ast $uropean

transformation countries. 1ince ++ the number of countries for which data is available

amounts constantly to +6/. ! table with more detailed information can be found in the

appendix.

Figure : *u(%er of !a(p&e Countries o)er $i(e

'he country sample consists of countries with uite different levels of development.

!ccording to the "#)4& (!N8 classification four types of countries are distinguishedA low,

+: or a description of the data see the website of the "#)4& (!N8 at A

H''


18/42+9

lower middle, upper middle and high income countries.+;

!s figure ; reveals, the country

sample consists of countries of all four classes, although because of data availability reasons

especially the low income countries are somewhat under and especially the high income

countries overrepresented.

Figure : Countr sa(p&e % or&d 3an4 c&assification

% &esults

In the following we give an overview on the most important and interesting results of

our aggregate capital stock estimations. &ue to space restrictions we concentrate on

reporting the estimation results for the absolute aggregate capital stocks, capital intensities

*capital per worker-, and capital coefficients *capital per unit of 7&


19/42+>

necessarily concentrate on subgroups of all sample countries. However, more detailed

results are summaried in the appendix.

%.1

A''re'ate capital stoc(s

In figure 9 we show a map visualiing the estimated aggregate stocks for 26+6.

1omewhat unsurprisingly, the countries with the most inhabitants tend to have also the

highest capital stocks, at least whenever they are at least upper middle income countries. In

figure > we show the 26 countries with the highest aggregate capital stocks in 26+6. In fact,

only three countries with less than 26 million inhabitants are among the 26 countries with

the largest capital stocksA the Netherlands, 1witerland and (elgium. 'he 0nited 1tates and

apan turn out to have by far the highest capital stocks. "hile %hina makes it to the third

place of the ranking, its capital stock is only slightly higher than one uarter of the capital

stock of the 0nited 1tates. 7ermany follows closely behind %hina. #n the fourth, fifth and

sixth place we find rance, the 0nited 8ingdom and Italy with only slightly differing capital

stocks. 'he next group of countries with similar aggregate capital stocks consists of 1pain,

%anada, 1outh 8orea, (rail, India, )ussia, 3exico and !ustralia. 'he final group is headed by

the Netherlands and includes 1witerland, !rgentina, 'urkey and (elgium.


20/42+=

Figure 6: 'sti(ated aggregate capita& stoc4s 2010+ 10/ countries

Figure 5: !a(p&e countries it7 7ig7est esti(ated aggregate capita& stoc4 2010

6

;666

+6666

+;666

26666

2;666

/6666

/;666


21/42+

Figure 8: Grot7 of esti(ated aggregate capita& stoc4s 1991-2010+ 10/ countries

In igure = we show a world map reporting the annual growth rates of the aggregate

capital stock in the sample countries in between ++ and 26+6.+9

It is easy to see that

capital growth varies significantly between our sample countries. In seven sample countries,

the capital stock decreased throughout the last two decades. !mong these countries are a

few !frican countries such as 7uinea *+.:=J-, Uambia *+.2:J-, 1wailand *6.:=J- and

7abon *6.:9J- but also %uba *+.;;J-. )ussia@s aggregate capital stock also decreased over

the last two decades by almost one percent per year. 'he worst development of the

aggregate capital stock of all sample countries occurred in the 0kraine *+.:J-.

igure reports the 26 sample countries with the highest aggregate capital stocks

growth rates in the last two decades. !erbaijan *+,6J- realied the highest annual growth

rate of the capital stock throughout the period of ++26+6. "ith an annual growth rate of

++,6J %hina follows on the second place. 'he other += countries realied annual growth

rates of the aggregate capital stock in between =.2J and ;.=J. 'his group of countries

+9"e chose the period of ++ to 26+6 because for this period data for all +6/ countries in our sample

are available.


22/4226

includes 9 !frican countries *1udan, 0ganda, 3oambiue, (otswana, 3adagascar and

'anania-, besides !erbaijan / additional transition countries *1lovenia, 4atvia and


23/422+

%hina overtook all other countries except apan and the 0nited 1tates. 'he only additional

change in positions occurred in 266: when the 0nited 8ingdom@s capital stock grew larger

than the one of Italy.

Figure 10: Gross fixed assets 1950-2010+ 10 countries it7 &argest aggregate capita& stoc4s

in 2010 "in ,!D of 2000#

#ver the period of ++26+6 the average aggregate capital stock of the +6/ sample

countries almost doubled from 9>9 bn. 01& in ++ to ++: bn. 01& in 26+6. However, this

increase in the mean level was not accompanied by a convergence of the capital stocks. #ver

the same horion, the standard deviation of the aggregate capital stocks rose strongly from

2+== bn. 01& in ++ to /==2 bn. 01& in 26+6.

%.2

"apital Intensity

"hile absolute aggregate capital stock data are often useful for empirical analyses

one might argue that the capital stock available per worker, i.e. capital intensity, is at least

from some perspectives the more interesting variable. High capital intensities indicate that

the amount of physical capital available per worker in the production process is also high.

6

;666

+6666

+;666

26666

2;666

/6666

/;666

+>6 +>; +=6 +=; +6 +; 2666 266; 26+6

0nited 1tates apan %hina 7ermany rance

0nited 8ingdom Italy 1pain %anada 8orea, )ep.


24/4222

In figure ++ we show a world map reporting capital intensities for the year 26+6. It is

easily visible that the ranking for this indicator is uite different from those reported in

section 9.+. $specially %hina, but also India and to some lower extent also (rail and )ussia

do not perform very well in terms of capital intensity. #n the other hand comparatively smallbut highly developed countries like the 1candinavian countries, Ireland, !ustria, 4uxemburg

and even the (ahamas appear among the 26 countries with the highest capital intensities.

Figure 11: Gross fixed assets per or4er 2010 "in ,!D of 2000#

!s figure +2 reveals, apan turns out to be the country with the highest capital

intensity, however, with only a small advantage before 4uxemburg. $ven 1witerland and

Norway exhibit considerably higher capital intensity than the 0nited 1tates. !lmost on the

same level as the 0nited 1tates we find countries like (elgium, Hong 8ong 1!), &enmark,

Iceland, !ustria, Ireland, inland, 7ermany, rance and Italy. 4agging slightly behind that

large group, the top 26 are completed by the Netherlands, !ustralia, the 0nited 8ingdom

and the (ahamas.


25/422/

Figure 12: Countries it7 7ig7est capita& intensities in 2010 "in ,!D of 2000#

#ver the period from ++ to 26+6, average capital intensity in our sample countries

rose from ;6:9+ to 9>969. However, again there dispersion within the sample also

increased. "hile the standard deviation of capital intensities in ++ was 9+=:: it rose to

=+9/: in 26+6. 'hus, we observe no convergence of capital intensities in the sample

countries.

%.3 "apital "oefficients

It is also an interesting uestion, how much capital a country needs to generate the

current output. In order to study this uestion, we calculate capital coefficients for all

countries in our country sample. 'he capital coefficient is simply the amount of capital

divided by the gross domestic product. 'he capital coefficient informs how much capital is

needed to generate one unit of output. igure +/ shows a world map with capital

coefficients. igure +: delivers an overview on the 26 countries with the highest capital

coefficients.

6

;6666

+66666

+;6666

266666

2;6666

/66666

/;6666

:66666


26/422:

Figure 1/: Capita& coefficients 2010 "in ,!D of 2000#

'he country with the by far highest capital coefficient is the 0kraine *=.6>-, followed

by 7abon *;.9=-, 4esotho *;.6+-, )ussia *:.-, 1wailand *:.>+-, $stonia *:./>-, apan *:.26-

and (runeiB&arussalam *:.62-. 'he following group of countries consists of the (ahamas,

.

'he countries with the lowest capital coefficients are 'ajikistan *+.>2-, the &ominican

)epublic *+.9;-, 3acao *+.:9- and 1udan *+.22-.

Interestingly enough, the mean capital coefficient of our sample countries remaineduite stable in between ++ and 26+6. It fell only slightly from /.+> in ++ to /.66 in 26+6.

3oreover, the capital coefficients show a strong tendency to converge, as the standard

deviation decreased from +.>; to 6./ throughout the last two decades.


27/422;

Figure 1: Countries it7 7ig7est capita& coefficients in 2010 "in ,!D of 2000

) Summary and "onclusions

'he lack of internationally comparable capital stock data has been a major obstacle

to empirical multicountry research on the role of physical capital in the process of economic

growth. In order to avoid this problem, various authors have constructed capital stock data

using some variant of the


28/4229

approaches used in the previous literature in order to avoid most of the problems of these

approaches.

'he resulting dataset is large enough to allow for pure cross section analyses as well

as for panel studies. !t least for the subsample of ;= countries, for which investment data

are available at least since +>6, the data can even be used to conduct timeseries analyses.

However, since for many countries *noncomparable- official aggregate capital stock data is

available, one might prefer the official data for the latter purpose.

'he database can be easily downloaded from our internet page. #ur approach allows

to extend the existing time series of capital stock estimations in a uite simple and

consistent way. 1ince the investment time series in the "orld &evelopment Indicators

database is updated regularly, we will extend the dataset in certain intervals to secure

availability of actual data.


29/422>

* &eferences

!nalysis, 0. &. *266/-. Fixed Assets and Consumer Durable Good in the United States,

1925-97"ashington, &%A 0.1. 7overnment /.

&erbyshire, ., 7ardiner, (., ? "aights, 1. *26+/-. $stimating the capital stock for the

N0'12 regions of the $02>.A..lied %&onomi&s, 5 )9*, pp. ++//++:.

uente, !. d., ? &omenech, ). *2666-. Human %apital in 7rowth )egressionsA How

3uch &ifference &oes &ata Suality 3akeD %&onomi&s De.artment /or0in a.er, 2(2,

3%CD, aris.

7riliches, U. *+=6-. )?& and the =-.


30/422=

1chmalwasser, #., ? 1chidlowski, 3. *2669-. 8apitalstockrechnung in &eutschland.

ol0s:irts&ha$tli&he Gesamtre&hnunen.

1chreyer,


31/422

Appendi+

$a%&e A-1: !tart and end of capita& stoc4 ti(e series % countr

Countr !tart 'nd

!lgeria +9= 26+6

!rgentina +96 26+6

!rmenia += 26+6

!ustralia +9: 26+6

!ustria +9 26+6

!erbaijan += 26+6

(ahamas, 'he +== 26+6

(angladesh +> 26+6

(elarus += 26+6

(elgium +9 26+6

(olivia +9 26+6

(otswana +>/ 26+6

(rail +9 26+6

(runei &arussalam +== 26+6

(ulgaria +> 26+6

%ameroon +>: 26+6

%anada +96 26+6

%ape Rerde +=; 26+6

%hile +96 26+6

%hina +9: 26+6

%osta )ica +96 26+6

%uba +9 26+6

%yprus +>: 26+6

%ech )epublic += 26+6

&enmark +9; 26+6

&ominican )epublic +96 26+6

$cuador +9: 26+6

$gypt, !rab )ep. +9: 26+6

$l 1alvador +96 26+6

$stonia +=> 26+6

$thiopia +=6 26+6

inland +96 26+6

rance +9 26+6

7abon +> 26+6

7ermany +9 26+6

7reece +96 26+6

7uatemala +96 26+6

7uinea +=; 26+6

Honduras +96 26+6


32/42/6

Hong 8ong 1!), %hina +9: 26+6

Hungary +96 26+6

Iceland +96 26+6

India +96 26+6

Indonesia +>= 26+6

Iran, Islamic )ep. +9: 26+6

Ireland +9 26+6

Italy +96 26+6

apan +96 26+6

ordan +>; 26+6

8aakhstan += 26+6

8enya +>= 26+6

8orea, )ep. +96 26+6

8yrgy )epublic += 26+6

4atvia += 26+64esotho +9 26+6

4uxembourg +96 26+6

3acao 1!), %hina +=+ 26+6

3acedonia, 5) += 26+6

3adagascar +=/ 26+6

3alaysia +96 26+6

3ali +>= 26+6

3alta +9 26+6

3auritius +>; 26+6

3exico +96 26+6

3oldova +6 26+6

3orocco +9; 26+6

3oambiue +> 26+6

Namibia +> 26+6

Netherlands +9 26+6

New Uealand +9 26+6

Nicaragua +96 26+6

Norway +96 26+6


33/42/+

1lovenia += 26+6

1pain +9 26+6

1udan +>; 26+6

1wailand +> 26+6

1weden +96 26+6

1witerland +96 26+6

1yrian !rab )epublic +>: 26+6

'ajikistan +=: 26+6

'anania += 26+6

'hailand +96 26+6

'unisia +96 26+6

'urkey +=9 26+6

0ganda +=+ 26+6

0kraine += 26+6

0nited 8ingdom +9 26+60nited 1tates +96 26+6

0ruguay +96 26+6

Reneuela, )( +96 26+6

Uambia +9 26+6

Figure A-1: Aggregate capita& stoc4s in &o inco(e countries "in %n ,!D of 2000#

+>+2

:2

+=>

;

/

2:

+>

/

+2

/>

6

26

:6

96

=6

+66

+26

+:6

+96

+=6

266


34/42/2

Figure A-2: Capita& intensities in &o inco(e countries "in ,!D of 2000#

Figure A-/: Capita& coefficients in &o inco(e countries "%ased on ,!D of 2000#

:2+=

/6/ /6/6

292

2:26

+=;

+=;9

+9>=

+/+6+2/=

9+

6

;66

+666

+;66

2666

2;66

/666

/;66

:666

:;66

:2+=

/6//6/6

292

2:26

+=;

+=;9

+9>=

+/+6+2/=

9+

6

;66

+666

+;66

2666

2;66

/666

/;66

:666

:;66


35/42//

Figure A-: Aggregate capita& stoc4s in &oer (idd&e inco(e countries "in %n ,!D of 2000#

Figure A-: Capita& intensities in &oer (idd&e inco(e countries "in ,!D of 2000#

>

+=

/=:

/>

/;>

+/ 2 /2 ;;

/2:

2; ; +=

9>6

2; :/

2+=

22 26

2/:

2=

6

;66

+666

+;66

2666

2;66

2=+

+:>>

+=29+=+9;

+;/6/+:;26

+2/;6

++/=6++6+:

/6

62; =9>+ =:+:=62=

9++/ ;>:6 ;:9:=2:

::;; :26> :+2=

2++

6

;666

+6666

+;666

26666

2;666

/6666


36/42/:

Figure A-6: Capita& coefficients in &oer (idd&e inco(e countries "%ased on ,!D of 2000#

Figure A-5: Aggregate capita& stoc4s in upper (idd&e inco(e countries "in %n ,!D of 2000#

=,6>

;,6+

:,>+

/,;/ /,:9

/,+; /,+2 /,6 /,6> /,69 /,6+

2,;+ 2,:: 2 ,:: 2,:2 2,/: 2,29 2,2/2,6: 2 ,6: 2,62

+,22

6

+

2

/

:

;

9

>

=

/9

+6+9

26:9

>9

/ 9/

:==2==

+

/=:

:/

26>+

2=+==

96+

2/ :+ +//

22:

;+ +: +: 9 = 99 +/: +;6

9+

229+

=692

/

6

+666

2666

/666

:666

;666

9666

>666

=666

666


37/42/;

Figure A-8: Capita& intensities in upper (idd&e inco(e countries "in ,!D of 2000#

Figure A-9: Capita& coefficients in upper (idd&e inco(e countries "%ased on ,!D of 2000#

>/9=:

;,9;9

:/;:>

:+>:9

:++9:

/,>;2

/,++,

/,69:

//2/,

//+:/

/6>6>

2,/>=

2=+;9

2>6:9

29;6=

29:+9

29:+6

2;;,/

2:/=6

2:2;,

2/;,2

222;:

2+,6>

2+>+=

+>:=,

+9::+

+;=,9

+;=>9

+;9+2

+:+=/

+6:,2

=,2>

6

+6666

26666

/6666

:6666

;6666

96666

>6666

=6666

7abon

!rgentina

3exico

'urkey

4atvia

0ruguay

Reneuela,)(

%hile

3auritius

3alaysia

ordan

)ussian2ederation

!lgeria

'unisia

Iran,Islamic)ep.

(otswana

+;>2/;

+;22>;

+;69=,

+:=::,

+::;+/ +

+/,,=

+69,;:

,>2:9

,:+,:

=:;9+

=2:2/

>/:=6

9+>;2

96::9

96/6,

;;=6

=

/

:,>+

6

;6666

+66666

+;6666

266666

2;6666

/66666

/;6666

:66666

apan

4uxembourg

1witerland

Norway

0nited1tates

(elgium

Hong8ong1!),%hina

&enmark

Iceland

!ustria

Ireland

2inland

1weden

7ermany

2rance

Italy

Netherlands

!ustralia

0nited8ingdom

(ahamas,'he

%anada

(runei&arussalam

1pain

7reece

8orea,)ep.

Documents

Estimating Aggregate Capital Stocks.pdf