Just how quickly has the Vietnamese capital stock grown over the past decade?

We use a detailed growth accounting method to estimate the annual growth rate of the aggregate Vietnamese capital stock over the period 1996-2005. Our results show that the rate was likely to have been in the vicinity of 7-8 % per annum. This is approximately 2 percentage points lower than estimates appearing in some previous studies.

1 Introduction

The Solow growth accounting equation (equation 1) is often used to estimate movements in total factor productivity (TFP)

A number of such studies have been undertaken for Vietnam (Tran Tho Dat 2004, Tang Van Khien

2005, Le Viet Anh 2006) A difficulty with applying (1) to the Vietnamese case is that, to date, the General Statistics Office (GSO) has not published

estimates of k Equation (1) thus presents the growth

accounting researcher with one equation, but two unknowns To solve this problem, researchers studying

the Vietnamese case must supply their own k

estimates These estimates tend to be high - in the vicinity of 10 per cent1 In this paper, we argue that these estimates may be too high, perhaps by as much

as 2-3 percentage points per annum

Just how quickly has the Vietnamese capital stock grown

over the past decade?

James A Giesecke and Tran Hoang Nhi

James A Giesecke and Tran Hoang Nhi, Centre of Policy Studies, Monash University, Australia

ABSTRACT

We use a detailed growth accounting method to estimate the annual growth rate of the aggregate Vietnamese capital stock over the period 1996-2005 Our results show that the rate was likely to have been in the vicinity of 7-8 % per annum This is approximately 2 percentage points lower than estimates appearing in some previous studies

Key words: Vietnamese capital growth rate, growth accounting, TFP growth.

FC

In such exercises, official statistical data on:

(a) percentage rates of change in real GDP at

factor cost ( ), employment (l), capital stocks (k) and land (n); and,

(b)shares of payments to labour (S L),

capital (S K ) and land (S N) in GDP at factor cost are

used to calculate movements in TFP (a).

FC R

gdp

where is the percentage change in the GDP at market prices deflator; is the percentage change in real GDP at market prices; is the percentage change in the GDP

at factor cost deflator; is the percentage change

in real GDP at factor cost; x i is the percentage

change in the quantity of activity i that is subject to indirect taxation; p iis the percentage change in the

price of activity i; t iis the percentage change in the

rate of indirect taxation of activity i)

For growth accounting, (6) is the key equation

We focus first on the indirect tax term, Movements in this term contribute to real GDP movements via allocative efficiency effects arising from the wedge that indirect taxes drive between values in use and values in supply for the activities

i that are indirectly taxed We do not have data on the billions of transactions that constitute i.

Instead, we simplify by defining three broad sets of activities that are the major bases for Vietnamese indirect taxation, and a residual The three broad activities are: (1) real private consumption, which

where:

is nominal GDP at market prices;

is nominal GDP at factor cost; and

is nominal collections of indirect tax revenue

We re-write (2) in terms of real variables and price indices as follows:

In this paper we provide annual estimates of the rate of growth of the aggregate Vietnamese capital stock over the period 1996-2005, using a detailed growth accounting framework Our approach takes into account not only changes in aggregate inputs of labour and land, but also the effects of movements of these factors between sectors of the economy Since GSO publishes data for real gross domestic production (GDP) at market prices, but not real GDP at factor cost, we must also account for movements in indirect taxes We argue that the growth rate of the aggregate capital stock can only be in the vicinity of 10% if the rate of TFP growth is implausibly low (in the vicinity of 0.0 to 0.5 % per annum) If, as we suspect, annual TFP growth has been about 1.5 % per annum, then the average capital growth rate over the past decade has been about 7.5% per annum

2 Methodology

We begin with equation (2), the definition of nominal GDP at market prices:

is the rate of indirect taxation on activity

i; and,

is the price of activity i

An example of activity i might be sales of imported five-seat-automobiles to households According to the tax code in 2005, such a transaction attracts tariff at the rate at 100%, special consumption tax (SCT) at the rate of 80% and value added tax (VAT) at the rate of 10% Dixon and Rimmer (2002, pp 187-188 and 230-232) show that if the levels variables in (3) are interpreted as Divisia indexes, then (3) and its components have the percentage change forms:

MP

GDP

FC

GDP

ITAX

where:

is real GDP at market prices;

is the GDP at market prices deflator;

is the GDP at factor cost deflator;

is the quantity of activity i that is subject

to indirect taxation;

MP R

GDP

MP GDP

P

FC GDP

P

i

X

i

T

i

P

MP MP MP FC FC FC GDP R GDP R i i i i i i

GDP p +gdp =GDP p +gdp +åX PT x p t+ + (4)

MP GDPFC FC ITAX i GDP GDPMP GDP GDPMP i ITAX i i

MP GDPFC FC ITAX i

R GDPMP R GDP i ITAX i

MP GDP

p

FC GDP

p

FC R

gdp

.

GDP i ITAX i

MP R

gdp

is subject to Special Consumption Tax (SCT) and much of the incidence of Value Added Tax (VAT);

(2) imports, which are subject to tariffs; and (3) natural resource extraction (mainly crude oil and gas), which is subject to natural resource tax

While this list represents only four taxes in over a dozen types of Vietnamese indirect taxation, they constitute a large share of total indirect taxation, starting at 63% of indirect tax revenue in

1996, rising to 96% by 2005 (see Table 3)

We define the percentage changes in the indirect

tax bases as follows: (1) for VAT and SCT, c R, the percentage change in real private consumption; (2)

for tariff revenue, m R, the percentage change

in import volumes; (3) for natural resource tax, , the percentage change in mining output;

and (4) for all other indirect taxes, , the percentage change in real GDP

On this basis, we assume that:

to use quantity weights Movements in l have been

defined as the percentage change in the number of persons This underestimates the growth in effective Vietnamese employment, by failing to account for movement of labour from low wage to

high wage activities Movements in k are either

assumed, or estimated via the perpetual inventory method However the perpetual inventory method implicitly weights movements in sector-specific capital using asset value weights This might over-estimate growth in the (rental weighted) capital stock, if Vietnamese investment is used inefficiently or directed towards sectors with relatively low rates of return2

The capital growth rate in any particular year can be calculated based on (8) as:

i VAT SCT TARF NRT Mining Other MP ITAX i ITAX R ITAX R ITAX R ITAX R

i S x S= + c S+ m S+ x +S gdp

Mining R

x

MP R

gdp

Substituting (1) and (7) into (6) provides:

MP GDPFC L K N

R GDPMP GDPFC GDPFC GDPFC ITAX VAT SCT TARF NRT Mining Other MP GDPMP ITAX R ITAX R ITAX R ITAX R

gdp S S l S k S n a

S S + c S m S x S gdp

= ë + + + +û

1 [ ].

.

MP GDPFC L GDPFC N GDPFC

R GDPMP GDPFC GDPMP GDPFC GDPMP GDPFC K ITAX VAT SCT TARF NRT Mining Other MP GDPMP GDPFC GDPMP ITAX R ITAX R ITAX R ITAX R

gdp S S l S S n S a k

S S S S + c S m S x S gdp

In equation (8), the percentage change

variables l, k and n, have precise meanings In particular, l is wage-bill weighted percentage changes in employment categories, k is rental

weighted percentage changes in types of physical capital, and n is rental weighted percentage changes in parcels of land Unfortunately, in implementing equations like (8) for Vietnamese growth accounting exercises, researchers have not always defined these variables using the appropriate factor payment weights Rather, they have tended

Our aim is o calculate k To do this using (9),

we need values for the year-on-year movements in the percentage change variables and annual values for the shares on the right-hand-side of the equation We explain in Section 3 our calculations

of these variables and shares

3 Data

3.1 Factor payment shares in GDP at factor cost ( , , )

Table 1 reports factor payment shares in GDP

at factor cost for the year 1996-2005 As we explain in Part 1 of the Appendix, we calculate factor shares for 1996, 2000, 2003, and 2005 using

IO and SAM data (GSO 1999, 2003, Jensen and Tarp 2007, Ministry of Finance (MOF) 2007), and GTAP 6.0 database (Dimaranan 2006) Values for intervening years are interpolated assuming linear growth

K GDPFC

GDPFC

S

L GDPFC

S

3.3 Indirect tax shares

Table 3 reports our estimates of the shares of aggregate indirect tax collections represented by

VAT and SCT ( ), tariffs ( ), natural resource tax ( ) and other indirect taxes ( ) We calculate these shares from IMF government revenue data (IMF 2003, 2006, 2007)

3.2 Factor payment and indirect tax shares in GDP at market prices ( and )

GSO does not publish income side data for GDP

at market prices Hence we must estimate and using other sources We do this by first calculating

the ratio of IMF estimates of Vietnamese indirect tax collections to our estimate of GDP at factor cost (IMF 2003, 2006, 2007) Then we calculate the shares

of GDP at factor cost and indirect tax collections in GDP at market prices Table 2 reports our share estimates More details on our approach are provided

in Part 2, Appendix

Source: Authors' calculations from GSO (1999, 2003), Jensen and Tarp (2007), MOF (2007) and GTAP

6.0 (Dimaranan 2006)

Table 1: Factor payment shares in GDP at factor cost, 1996-2005

ITAX GDPMP

S

GDPFC GDPMP

S

ITAX GDPMP

S

NRT ITAX

S

VAT SCT ITAX

S +

GDPFC GDPMP

S

TARF ITAX

S

Other ITAX

S

Table 2: Income-side GDP shares

Source: Authors' calculations from GSO (1999, 2003), Jensen and Tarp (2007), MOF (2007), GTAP 6.0

(Dimaranan 2006), and IMF (2003, 2006, 2007)

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Average

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Average

1 Labour ( )

2 Capital ( )

3 Land ( )

4 Total

0.56 0.32 0.12

1.00

0.56 0.32 0.12

1.00

0.55 0.33 0.12

1.00

0.55 0.33 0.12

1.00

0.55 0.34 0.12

1.00

0.55 0.34 0.12

1.00

0.54 0.34 0.12

1.00

0.54 0.34 0.12

1.00

0.54 0.35 0.11

1.00

0.54 0.35 0.11

1.00

0.55 0.34 0.12

1.00

L GDPFC

S

K GDPFC

S

N GDPFC

S

GDPFC GDPMP

S

ITAX GDPMP

S

1 Share of GDPFC in GDPMP ( )

2 Share of ITAX in GDPMP ( )

3 Total

0.85 0.15

1.00

0.87 0.13

1.00

0.87 0.13

1.00

0.88 0.12

1.00

0.89 0.11

1.00

0.88 0.12

1.00

0.88 0.12

1.00

0.87 0.13

1.00

0.87 0.13

1.00

0.87 0.13

1.00

0.87 0.13

1.00

Notes: (a): Turnover tax prior to 1999.

Source: Calculated from IMF (2003, 2006, 2007).

VAT ITAX

S

SCT ITAX

S

Other ITAX

S

NRT ITAX

S

TARF ITAX

S

Mining R

x

MP R

R

R

x

Table 3: Shares of important taxes in total indirect tax revenues at current prices

3.4 Real GDP at market prices, real private consumption, import volumes, mining output ( , cR, mR, and )

The GSO publishes annual estimates of real GDP at market prices, real private consumption, aggregate imports at constant prices, and mining

output at constant prices (GSO 2007) We calculate the annual percentage changes in these series,

allowing us to evaluate , c R , m R, and Values for these variables are reported in rows

1 - 4 respectively of Table 4

Table 4: Real GDP, private consumption, import volume and output

of mining sector at market prices

(percentage change from previous year)

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Average

1 Real GDP ( )

2 Real private

consumption (c R)

3 Import volume

(m R)

4 Output of mining sector ( )

5.Weighted average indirect taxes

9.3 9.1 15.8 14.7

11.1

8.2 5.9 16.8 14.7

9.9

5.8 4.5 31.6 15.3

12.4

4.8 2.6 22.0 16.4

9.0

6.8 3.1 31.7 11.2

11.2

6.9 4.5 13.3 6.4

7.3

7.1 7.6 -3.8 4.2

4.3

7.3 8.0 -0.8 8.0

5.9

7.8 7.1 14.9 14.4

10.1

8.4 7.3 47.0 2.4

13.9

7.23 5.94 18.0 10.7

9.6

MP R

gdp

Mining R

x

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Average

1 VAT(a) ( )

2 Special consumption tax ( )

3 Natural Resource tax ( )

4 Tariff ( )

5 Other taxes ( )

Total

0.25 0.08 0.08 0.23 0.37

1.00

0.30 0.12 0.09 0.25 0.25

1.00

0.27 0.13 0.07 0.25 0.28

1.00

0.36 0.09 0.10 0.24 0.22

1.00

0.35 0.11 0.15 0.22 0.17

1.00

0.34 0.11 0.15 0.25 0.15

1.00

0.39 0.11 0.13 0.24 0.13

1.00

0.42 0.12 0.14 0.24 0.09

1.00

0.41 0.13 0.18 0.21 0.06

1.00

0.41 0.14 0.20 0.19 0.07

1.00

0.35 0.11 0.13 0.23 0.18

1.00

3.5 Economy-wide growth in labour

and land inputs (l and n)

GSO does not publish estimates of growth in

wage-bill-weighted employment (l) and rental-weighted land use (n) We calculate l and n using

data from GSO (2007), Jensen and Tarp (2007) and

GTAP 6.0 (Dimaranan 2006) Our estimates for l and n are reported in row 11, Table 6 and row 9,

Table 8 We explain our method for calculating these values below

3.5.1 Wage-bill-weighted employment growth (l)

Equation (10) defines l, the percentage change in

wagebill-weighted employment growth:

where is the share of payments to labour in

industry j in economy-wide payments to labour, and l j is the percentage change in employment in

industry j Hence, to calculate l, we require estimates of and l j Tables 5 and 6 report our estimates To calculate we use industry data on payments to labour obtained from GSO (1999, 2003), Jensen and Tarp (2007), MOF (2007) and GTAP 6.0 (Dimaranan 2006)

j

L j j

j L

S

j L

S

Table 5: Wage-bills shares, by industry ( )

Source: Authors' calculations from GSO (1999, 2003), Jensen and Tarp (2007), MOF (2007) and GTAP

6.0 (Dimaranan 2006)

We elaborate on our use of these data in calculating in Part 1(b) of the Appendix Values

for l j are calculated from GSO employment data for nine broad sectors (GSO 2005, 2007) Row 11

of Table 6 reports our estimates for l using data

from Table 5 and 6 to implement equation (10)

Note that l is substantially higher than the simple

growth rate in total number of persons employed (row 10 Table 6) This reflects inter-sectoral movements of labour, particularly movements from low-paid industries such as agriculture, to high-paid industries such as manufacturing

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Average

1 Agriculture and forestry

2 Fishing

3 Industry

4 Construction

5 Trade

6 Hotels, restaurants

7 Transport, communications

8 Culture, health, education

9 Other services

10 Total

0.29 0.03 0.19 0.06 0.10 0.03 0.03 0.08 0.19

1.00

0.27 0.03 0.20 0.06 0.09 0.03 0.03 0.08 0.20

0.99

0.27 0.03 0.21 0.06 0.09 0.03 0.03 0.08 0.21

1.00

0.25 0.03 0.22 0.05 0.08 0.03 0.03 0.08 0.23

1.00

0.23 0.03 0.25 0.05 0.11 0.03 0.04 0.09 0.16

1.00

0.23 0.04 0.27 0.06 0.10 0.03 0.04 0.09 0.15

1.00

0.22 0.04 0.29 0.06 0.09 0.03 0.04 0.09 0.15

1.00

0.21 0.04 0.31 0.06 0.08 0.03 0.04 0.09 0.15

1.00

0.20 0.04 0.31 0.06 0.08 0.03 0.04 0.09 0.15

1.00

0.20 0.04 0.31 0.07 0.08 0.04 0.04 0.09 0.15

1.00

0.24 0.03 0.26 0.06 0.09 0.03 0.04 0.08 0.18

1.00

j L

S

j L

S

j L

S

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Average

1 Agriculture and forestry

2 Fishing

3 Industry

4 Construction

5 Trade

6 Hotels, restaurants

7 Transport, communications

8 Culture, health, education

9 Other services

10 Total employment

11 Wagebill weighted

employment (l)

1.3 10.1 3.5 3.4 7.0 5.0 4.1 2.1 3.1

2.2 3.3

1.2 9.5 3.4 3.5 7.0 5.4 4.1 2.1 3.2

2.2 3.2

1.1 7.9 3.4 3.5 7.1 5.4 4.1 2.0 3.4

2.1 3.2

1.0 9.4 3.5 3.4 7.0 5.5 4.1 2.0 3.5

2.1 3.2

0.8 9.1 3.4 3.3 7.0 5.5 3.9 2.0 3.6

2.0 3.3

0.2 9.3 6.7 19.0 5.6 3.8 5.4 4.6 6.4

2.6 5.5

-0.9 18.4 7.0 18.2 5.4 2.2 0.3 5.7 9.0

2.5 6.0

-0.2 3.4 9.3 10.6 5.9 3.4 1.0 5.8 9.2

2.7 6.1

-0.4 5.9 6.2 13.9 5.2 2.1 0.7 4.6 10.5

2.5 5.4

-0.7 5.5 8.5 3.9 3.5 1.6 -4.5 4.1 20.1

2.3 6.4

0.3 8.8 5.5 8.1 6.0 4.0 2.3 3.5 7.1

2.3 4.6

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Average

1 Paddy

2 Raw rubber

3 Coffee bean

4 Sugar cane

5 Other crops

6 Forest

7 Fish farming

8 Total

0.51 0.01 0.02 0.03 0.20 0.02 0.21

1.00

0.51 0.01 0.02 0.03 0.22 0.02 0.18

1.00

0.52 0.02 0.02 0.03 0.23 0.02 0.16

1.00

0.52 0.02 0.03 0.02 0.25 0.02 0.15

1.00

0.52 0.02 0.03 0.02 0.26 0.02 0.13

1.00

0.49 0.02 0.03 0.02 0.26 0.02 0.15

1.00

0.47 0.02 0.03 0.02 0.25 0.02 0.19

1.00

0.46 0.02 0.02 0.02 0.24 0.02 0.22

1.00

0.46 0.02 0.02 0.02 0.24 0.02 0.21

1.00

0.47 0.02 0.02 0.02 0.25 0.02 0.20

1.00

0.49 0.02 0.02 0.02 0.24 0.02 0.18

1.00

3.5.2 Rental-weighted growth of land use areas (n)

Equation (11) defines n, the percentage change in rental-weighted land-use:

industry j in economy-wide payments to land, and

n j is the percentage change in land use in industry

j Clearly, to calculate n, we require estimates

of and n j Tables 7 and 8 report our estimates

To calculate we use industry data on payments

to land calculated from GSO (1999, 2003), Jensen and Tarp (2007), MOF (2007) and GTAP 6.0 (Dimaranan 2006)

where is the share of payments to land in

Table 6: Number of employed persons (percentage change on previous year) (l j )

Source: Employment data in rows 1 - 10 are from GSO (2005, 2007) We calculate the weighted change

in employment at row 11 by computing shares from rows 1-9 and multiplying by corresponding percentage changes in rows 1 - 9 of Table 5 (see text for details)

j

N j j

j N

S

j N

S

j N

S

j N

S

Table 7: Shares in land rentals, by industry ( )

Source: Authors' calculations from GSO (1999, 2003), Jensen and Tarp (2007), MOF (2007) and GTAP

6.0 (Dimaranan 2006)

3.6 Growth in total factor productivity (a)

Recall that to calculate k, an independent estimate of a is required We take our preferred estimate for a (1.5 per cent) by placing it at the

low-end of the range of findings from previous studies (1.7 - 3.4 %)

Many studies on Vietnam's economic progress over the last decade have attributed the economy's strong growth to high rates of technical progress (see, for example, Van Arkadie and Mallon 2003, Tran Nam Binh and Pham Do Chi

2003, Tran Tho Dat 2004, Tran Hoang Nhi 2007,

Vu Quoc Ngu 2002) This is consistent with the strong focus of policy on microeconomic reform

Over the past decade, such policies have been directed

at state enterprise reform, public administration reform, private sector development, further removal of barriers to international trade, and encouragement of foreign direct investment These developments have improved the functioning of market mechanisms, increased domestic competition and competition for international markets, and increased productive and allocative efficiency High levels of FDI are also likely to have contributed to productivity growth via diffusion of foreign technology and management practices Previous studies have estimated Vietnam's economy-wide TFP growth over the past decade to have been in the range of 1.7 to 3.4% per annum (Tran Tho Dat 2004, Le Viet Anh 2006, Tran Hoang Nhi 2007) Sectoral studies have also estimated high rates of TFP growth For example,

We elaborate on our use of these data in calculating in Parts 1(a) and 1(c) of the

Appendix Values for n j are calculated from GSO data on the area of land used in agriculture, forestry and fishing (GSO 2005, 2007) Row 9 of Table 8

reports our estimates for n using data from Table 7

and 8 to implement equation (11) On average, the growth in n is slightly higher than the simple average growth rate in total area used in agriculture (row 8 Table 8) This reflects above-average growth in fish farming and below-average growth

in paddy

j N

S

Table 8: Land used in agriculture, forestry and fisheries (percentage change from previous year)(n j)

Source: Land rental data in rows 1 - 7 are from GSO (2005, 2007) We calculate the rent-weighted

change in land area at row 9 by computing shares from rows 1-7 and multiplying by corresponding percentage changes in rows 1 - 7 of Table 6 (see text for details)

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Average

1 Paddy

2 Raw rubber

3 Coffee bean

4 Sugar cane

5 Other crops

6 Forest

7 Fish farming

8 Total

9 Rental-weighted land

areas (n)

3.5 -8.7 36.4 5.4 4.5 -3.2 9.9

4.2 5.6

1.4 36.7 33.9 8.4 2.9 9.3 1.1

3.5 3.2

3.7 9.9 8.9 10.1 2.1 -6.0 4.0

3.6 3.6

4.0 3.4 28.9 21.6 3.2 10.3 0.0

4.8 4.4

0.2 4.3 17.6 -12.2 7.3 -14.6 22.4

3.1 4.9

-2.3 0.9 0.6 -3.8 1.1 -2.9 17.7

-0.2 1.8

0.2 3.1 -7.6 10.1 8.4 -0.4 5.6

2.7 3.3

-0.7 2.8 -2.3 -2.1 5.2 -4.6 8.8

1.5 2.8

-0.1 3.0 -2.6 -8.7 5.5 1.7 6.1

1.8 2.4

-1.6 6.3 0.1 -6.9 4.6 -3.9 3.5

0.9 1.0

0.8 5.7 10.3 1.7 4.5 -1.7 7.7

2.6 3.3

Tang Van Khien (2005) estimated annual TFP growth in manufacturing to be 1.6% over

1996-2003 Vu Quoc Ngu (2002) annual TFP growth in state owned enterprises to be 5.37% over the period 1990-1998 For our calculations, we adopt a conservative value of 1.5 % as our preferred esti-mate for TFP growth However we also provide

Table 10 to show the sensitivity of our k estimate to different values for a

4 Estimated results

Table 9 brings together our estimates for the

right-hand-side terms of Equation (9) Row 6 of

Table 9 reports the resulting estimates for k The

path of these estimates appears plausible: in 1996 and 1997 we see high capital growth rates, reflecting high levels of FDI; over 1998 and 1999 we see low levels of capital growth as this FDI dried up in the face of the Asian financial crisis; consistent with observations that Vietnam recovered quickly from this crisis3, a relatively stable capital growth rate of approximately 7% per annum is established from

2000 to 2005 The annual average capital growth rate over the whole period is 7.5%

Table 9: Right-hand-side components of equation (9)

Since there is uncertainty about the value of TFP growth, we undertake sensitivity analysis

around the value of a Table 10 reports the results of applying equation (9) using alternative values for a.

Table 10 shows that the high capital growth rates that have appeared in some previous studies, that is annual capital growth in the range of 9-11%, imply rates of TFP growth that are implausibly low

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Average

1 Real GDP ( )

2 Wage-weighted

employment (l)

3 Rental-weighted land

use (n)

4 Weighted indirect taxes ( )

5 Total factor

productivity (a)

6 Share of GDPFC in GDPMP ( )

7 Labour share in GDPFC ( )

8 Capital share in GDPFC ( )

9 Land share in GDPFC ( )

10 Hence, growth rate

of capital stock (k)

9.34 3.28 5.56 11.1 1.50 0.85 0.56 0.32 0.12

15.8

8.15 3.20 3.19 9.94 1.50 0.87 0.56 0.32 0.12

13.1

5.76 3.20 3.60 12.4 1.50 0.87 0.55 0.33 0.12

3.29

4.77 3.24 4.39 8.98 1.50 0.88 0.55 0.33 0.12

1.10

6.79 3.35 4.91 11.2 1.50 0.89 0.55 0.34 0.12

6.86

6.89 5.52 1.78 7.31 1.50 0.88 0.55 0.34 0.12

6.26

7.08 6.02 3.29 4.33 1.50 0.88 0.54 0.34 0.12

6.81

7.34 6.12 2.75 5.85 1.50 0.87 0.54 0.34 0.12

7.09

7.79 5.42 2.45 10.1 1.50 0.87 0.54 0.35 0.11

7.84

8.44 6.44 1.05 13.9 1.50 0.87 0.54 0.35 0.11

7.12

7.23 4.57 3.29 9.49 1.50 0.87 0.55 0.34 0.12

7.45

L GDPFC

S

N GDPFC

S

K GDPFC

S

GDPFC GDPMP

S

MP R

gdp

.

S åS x

Our capital growth estimates are approximately 2-3 percentage points lower than those in some previous studies Our estimates are lower for two main reasons:

1 Application of the growth accounting framework requires inputs of wage-bill-weighted movements in employment, not total hours The former has grown more quickly than the latter as labour has moved from agriculture to manufacturing

2 We take into account changes in indirect taxes As evident from rows 1 and 4 of Table 9, real indirect taxes have grown at a faster rate than real GDP at market prices for most years of the study period Hence, for most years, growth in real GDP

at factor cost has been lower than growth in real GDP at market prices This presents an issue when researchers use GSO estimates of growth in real GDP at market prices to inform values for real GDP at factor cost ( ) in growth accounting applications of equation (1)

5 Conclusion

Growth accounting equations like (1) are typically used to calculate TFP growth from published data on such variables as growth in real GDP, employment and capital stocks A difficulty

in applying such an equation to Vietnam is the absence of official data on growth in capital stocks

Hence researchers applying (1) to estimate TFP

have needed to produce independent estimates for k.

These estimates have been in the range of 8.5 - 11 %

Our paper suggests that these estimates may have been too high, and that the true figure is probably in the vicinity of 7.5 % for the period 1996-2005

The absence of official capital stock data is a result of the more general limitation that the GSO does not publish yearly supply-side estimates of the components of real GDP at market prices As

we discussed in Sections 3 and in the Appendix, this also presented challenges for our calculations, which we overcame by going to other data sources such as Jensen and Tarp (2007), IMF (2003, 2006, 2007) and GTAP (Dimaranan 2006) Future GSO estimates of the components of income-side GDP will be very valuable This would greatly assist researchers interested in studying supply-side structural features of Vietnam's economic development

Appendixes

1 Calculation of factor payments for the period 1996-2005

There are four main sources of Vietnamese sectoral-level factor payment data for the study period They include three sets of input-output (IO) data for the years 1996, 2000, 2003 and 2005 (GSO

1999, 2003, MOF 2007), and a social accounting matrix (SAM) for 2003 (Jensen and Tarp 2007) Of these, only the SAM has payment data for all three factors of production: labour, capital and land The input-output tables contain only data on

Table 10: Sensitivity analysis: Capital growth rates under alternative TFP growth assumptions

2.0 6.0

1.7 6.9

1.5 7.5

1.0 8.9

0.5 10.4

0.0 11.9

2.5 4.5

TFP growth rate (a)

Annual average (1996-2005)

capital growth rate (k)

Source: GSO (2005, 2007).

FC R

gdp