Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
Preliminary.
Economic Reforms and their Impact on the Indian
Manufacturing Sector: Competitiveness, Exports and Employment
Eckhard Siggel, Department of Economics, Concordia University, Montreal e-mail: [email protected] This paper has been prepared for the 14th World Congress of the International Economic Association in 2005, Morocco. Research assistance was provided by Jade Haddad.
Introduction
The reforms of the Indian economic system are wide-ranging in terms of policy areas and
sectors targeted. In this paper we focus on the large-scale manufacturing sector as
recorded by the Annual Survey of Industries (ASI). The policy areas that affect this
sector most strongly, by creating incentives and disincentives, are policies regulating
imports and exports, the exchange rate, the interest rate, as well as changes in the
regulatory framework directed towards industries. Unfortunately not all of these reform
components can be quantified. Therefore it is difficult to unambiguously attribute the
changes in growth of value added, employment and exports to changes of specific policy
variables. An attempt is made, however, to isolate some of the observable changes and to
relate them statistically to the measurable changes in policy variables.
Our approach is to use the available quantitative evidence of changes in trade
policy (rates of protection), in exchange rate policy (the real exchange rate), in interest
rate policy (the market and shadow rates of interest), together with output and cost data
from the ASI, during the study period (1987/88 to 1997/98), and to relate observable
changes in competitiveness, comparative advantage, exports and other variables to the
changes of these policy variables. The definition of the study period as well as the choice
of manufacturing as representative of the modern sector, were dictated by the availability
of the data for the chosen method of analysis.
We concentrate on the reforms undertaken in the aftermath of the 1991 crisis and
their subsequent effects on the performance of the manufacturing sector. There is an
ongoing debate in the literature (as well as in political circles) as to whether these effects
can be attributed to reforms implemented in the pre-1991 era and if the reform movement
as a whole actually started with those initial steps taken throughout the 1980s. Rodrik
(2002) argues that tentative measures taken under the Rajiv Ghandi government of the
1980s led to disproportionately high growth, while the reforms undertaken in and after
1991 had a far smaller impact with respect to GDP growth. Panagariya (2004) refutes
this argument and argues in favour of the 1990s reforms stating: “Growth during the
1980s was fragile, highly variable from year to year, and unsustainable. In contrast, once
2
the 1991 reforms took root, growth became less variable and more sustainable with even
a slight upward shift in the mean growth rate” (Panagariya, 2004, p. 5).
Panagariya goes on to argue that despite the 1980s reforms’ limitations in terms
of scope and vision, they differed markedly from the “isolated and sporadic”
liberalization measures implemented throughout the 1960s and 1970s, and can therefore
be seen as forerunners to more “systematic and systemic” reforms of the 1990s. In
providing this argument, Panagariya draws on support from Ahluwalia (2002, p. 67), who
stated that “while the growth record in the 1990s was only slightly better than that in the
1980s, the 1980s growth was unsustainable, fuelled by a build-up of external debt that
culminated in the crisis of 1991.”
Srinivasan and Tendulkar (2003) provide an export-oriented view of the reforms
undertaken in the 1980s by stating that the increase in Indian exports over the 1980s
reform era was due (mostly) to an exchange rate depreciation attributed more to
exogenous forces than to “explicit policy reforms aimed at reducing the trade barrier”.
1 Reform incidence: a view from the manufacturing sector
First, we provide evidence of the reform incidence as it affects the industrial sector. In the
area of trade and industrial policies we record the rates of nominal and effective
protection as well as their changes over the study period. Similar observations are made
later about foreign exchange and interest rate policies.
1.1 Trade and industrial policy reforms
In the area of trade and industrial policies, the reforms included first the elimination of
quantitative restrictions, which had formerly affected most industries. However their
elimination was not completed until the year 2001. This policy change meant that
protection shifted entirely, but gradually, to the import tariff. The second important
change is the reduction of tariff rates and the resulting change in the structure of
3
protection. Both policy interventions were accompanied by a host of other changes in the
regulatory framework and in particular the industrial licensing regime.
In order to measure the incidence of such changes it is necessary to observe the
price changes of domestic prices relative to the corresponding free-trade or border prices.
This is necessary because the main effect of quantitative restrictions, such as quotas, on
the incentive structure is to raise the domestic price relative to the free-trade price. This is
a very difficult task and has been undertaken systematically only in a few countries and
by few governments or expert agencies.1 A recent study by Pandey and associates (1998)
of price-based rates of protection is available for India and has been used here for the
purpose of quantifying the trade reform incidence as well as our computations of
competitiveness indices.
Table 1 : Nominal (NRP) and effective (ERP) rates of protection in manufacturing 1987/88 1997/98 Change in1
Industry NRP ERP NRP ERP NRP ERP Food products 1.3549 1.4666 0.3596 0.3909 0.2654 0.2666 Bever.& tobac. 1.5000 1.6318 0.5985 0.9288 0.3990 0.5692 Textile industry 1.3549 1.5168 0.3753 0.4239 0.2770 0.2795 Garments 1.4093 1.3931 0.4635 0.5141 0.3289 0.3690 Wood products 1.3530 1.4395 0.4500 0.5782 0.3326 0.4016 Paper & printing 1.5979 1.6123 0.3465 0.3627 0.2168 0.2250 Leather products 1.4500 1.4641 0.3854 0.4223 0.2658 0.2884 Rubber, plastic, petrol. & coal 1.4123 1.5151 0.3809 0.5027 0.2697 0.3318 Chemicals 2.0156 2.5922 0.3532 0.3827 0.1752 0.1477 Non-met. min. 1.4561 1.5144 0.4217 0.5134 0.2896 0.3390 Basic metals 1.7456 2.5916 0.3249 0.3674 0.1861 0.1418 Metal products 1.6215 1.5641 0.3519 0.3740 0.2170 0.2391 Machinery 1.4148 1.3660 0.3367 0.3362 0.2380 0.2461 Transport equip. 1.2996 1.1903 0.3927 0.4186 0.3022 0.3517 Other man. ind. 1.4905 1.4780 0.3933 0.4081 0.2639 0.2761 Average 1.5096 1.6088 0.3976 0.4678 0.2634 0.2908 Notes : 1 NRP97/98/NRP87/88 and ERP97/98/ERP87/88
Source : Pandey et al.(1998)
1 The World Bank has undertaken studies of nominal and effective rates of protection based on price controls in various countries in the 1970s and 1980s. One of the few countries, in which government agencies undertook such studies is Mexico (ten Kate, 1985)
4
Based on this study the following picture emerges of the structure of protection
and its changes between 1987/88 and 1997/98. Table 1 shows that for manufacturing as a
whole nominal protection declined from 151 percent in 1987/88 to 39.8 percent in
1997/98, a reduction of about 74 percent. Effective protection declined from 160.9
percent in 1987/88 to 46.8 percent, a reduction of about 71 percent. The reduction of
protection was most dramatic for the chemical and basic metal industries: 72 and 74
percent, respectively, for NRP and ERP.
Panagariya (2004) provides us with a more detailed perspective on the incidence
of tariff reductions undertaken after the June 1991 crisis. According to WTO (1998), as
surveyed by him, the import-weighted average tariff rate stood at 87 percent throughout
the 1990–1991 period, with the highest at 355 percent. Tariff reform was undertaken
through effective reductions in the number of tariff bands and a consistent compression
of the top tariff rate falling to 85 percent in the 1993–94 period, 50 percent in the 1995–
96 period and finally 25 percent in the 2003–04 period.
As far as regulatory policies for the manufacturing sector are concerned,
Panagariya (2004) outlines the effectiveness with which the July 1991 “Statement of
Industrial Policy” sought to (and did) eliminate investment licensing and entry
restrictions on companies under the purview of the Monopolies and Restrictive Trade
Practices (MRTP) act. Throughout the 1990s reform era, and following the July 1991
“Statement of Industrial Policy”, all investment licensing (irrespective of amount) was
abolished across all sectors except 18 (outlined in Annex II of the policy statement and
later reduced to five), the public sector monopoly was limited to eight sectors (listed in
Annex I, and selected according to security and strategic considerations), and pre-entry
inspection of investment decisions of MRTP companies along with provisions relating to
mergers, takeovers, and amalgamations were repealed.
5
1.2 Foreign exchange policy
Perhaps the most powerful policy instrument with regard to the incentive regime is the
exchange rate. Countries adhering to a fixed exchange rate regime have a tendency to
function with a misaligned exchange rate. Fixed exchange rates become misaligned when
the fixed rate is not periodically adjusted to the differential between domestic and foreign
rates of inflation. Trade liberalization often goes in tandem with liberalization of the
exchange rate, which implies letting the price of foreign currencies be determined by
supply and demand. The Indian reforms of 1991 and consecutive years also included such
a re-alignment of the rupee.
The July 1991 reform package included a devaluation of the rupee by 22 percent
against the dollar, driving it from Rs 21.2 to Rs 25.8 per dollar. Following this
devaluation, the government introduced a dual-exchange rate system in February of 1992,
allowing importers to fully operate on the open foreign exchange market while exporters
were authorized to sell 60 percent of their foreign exchange at open market prices while
the other 40 percent were sold at the lower official price (Panagariya, 2004). Within one
year the exchange rate was unified.
The degree of misalignment can be computed using the real exchange rate. In
other words, rather than observing a shadow exchange rate and the divergence of the
market rate from the shadow rate, one infers the degree of misalignment by observing
domestic and foreign price changes, by computing the real exchange rate over time and
by identifying a base year in which the misalignment was known to be minimal.
The shadow exchange rate and the implicit rate of currency overvaluation are
estimated here using this method based on the real exchange rate. It is assumed that the
year 1994, in which the exchange rate was unified and the rupee was made fully
convertible on the trade account, was a year of minimal misalignment; it is taken as
benchmark year, in which the real exchange rate index equals 100. The real exchange rate
depreciated between 32 and 53 percent between 1987/88 and 1994, depending on which
price index is chosen2. This, together with the assumption of zero misalignment in 1994,
2 Using wholesale prices, for which the inflation differential is largest, the real exchange rate depreciated by 32 percent, while using the GDP deflator leads to real depreciation of 53 percent. Using consumer prices,
6
implies that in 1987/88 the rupee was overvalued between 48 and 113 percent. We are
using an intermediate rate of 60 percent as the rate of currency overvaluation in 1987/88,
which is close to the rate based on consumer prices.
One could argue that this rate may overstate the real overvaluation because the
exchange rate may have overshot its target in 1994 and may have been undervalued after
several years of strong nominal depreciation. This is unlikely, however, because the
rupee continued to depreciate in real terms in 1995; it appreciated somewhat in 1996, but
depreciated again in 1997 and 1998, returning to its benchmark value of 100 in the latter
year.
Using the real exchange rate based on the hypothesis of well-alignment in 1994,
the rupee was then roughly at par in 1997/98. Using wholesale prices, it may even have
been slightly undervalued at 4 percent. The elimination of currency overvaluation implies
that Indian industries became more competitive by this aspect of external policy reform.
1.3 The price of capital and foreign investment
One of the main targets of policy reform is the price of capital and the access to foreign
capital markets. To the extent that the capital account is liberated, the price of capital is
increasingly determined by the international interest rate, such as the LIBOR. Capital
account liberalization is usually the last step in financial reforms. In the earlier stages of
reform, the price of capital remains a domestic variable and tends to be influenced mainly
by the state of the financial sector, the degree of financial repression and the interest rate
policies of governments.
India’s reforms have included financial reforms, which have had the double effect
of reducing the domestic cost of capital and opening the country to foreign capital
inflows. The domestic price of capital is measured here by the lending rate, which was
lowered from 16.5 percent in 1987/88 to 13.8 percent in 1997/98. The shadow price of
capital is computed here as the LIBOR adjusted for the inflation differential between
the real depreciation was 40 percent. The exchange rate used in this calculation is the one of rupees per SDR, and the foreign price indices relate to the industrial country aggregate as reported by the IMF (IFS, current issues).
7
India and OECD countries. It remained fairly stable at around 11 percent (11.5 percent in
1987/88 and 11.3 percent in 1997/98. Therefore, the interest rate premium paid by Indian
investors has been reduced by 2.5 percent, and this reduction may considered as an
indicator of financial liberalization.
More important than the cost of capital, however, may be the influx of foreign
investment, which also occurred under the reforms. Its potentially great importance stems
from the fact that it comes bundled with foreign technology whenever it is direct and
long-term investment. In 1987/88 foreign direct investment was literally non-existent in
India, but it started to flow in by 1991 and reached a total of about three billion dollars in
1997/98. Unfortunately, we have no information on the amount of foreign direct
investment received by each industry in the manufacturing sector.
The 1991 reform package called for abolishing the 40 percent threshold on
foreign direct investment, and empowered the Reserve Bank of India to approve equity
investments of up to 51 percent in 34 industries through the development of an automatic
approval concept, Panagariya (2004)3.
2 Changes in competitiveness and comparative advantage
The reduction of protection observed during the study period has certainly had the effect
of increased competition from foreign imports, which in turn induced firms to cut costs.
Such cost reductions can result from shedding redundant workers, by adopting new
production processes, which usually require new investments, and/or by increasing the
efficiency of capital and other input use. Successful cost reductions should then result in
greater competitiveness, both in the domestic market, vis-à-vis imports, and in export
markets. The purpose of this section is to examine how the competitiveness of industries
has changed during the study period. It is of particular interest to see whether the
observed changes have also led to an increase of comparative advantage.
3 A comprehensive list of the concerned industries is listen in Annex 3 of the July 24, 1991 “Statement of Industrial Policy”, Government of India
8
2.1 The measurement of competitiveness
Competitiveness is measured here by three unit cost ratios developed and applied in
several earlier studies (Siggel, Ssemogerere, 2004; Cockburn et al., 1999), including one
on India in the 1980s (Siggel, 2001)4. The first ratio, UCd, which we consider to be an
indicator of domestic competitiveness, is the ratio of unit costs to the output price, both
measured in domestic, possibly distorted, prices. Since unit costs differ from the domestic
price only by excess profits, this indicator is simply a measure of profitability in the
protected domestic market. It differs, however, slightly from the rate of return, because
own capital enters into the unit costs with its opportunity costs at market prices.
The second unit cost ratio, UCx, is an indicator of export competitiveness. It
divides the full unit cost at domestic prices by the free-trade price. This indicator tends to
regularly exceed unity because the unit costs include tariff margins of all traded
intermediate inputs, while the output price does not.
The third unit cost ratio, UCs, is an indicator of comparative advantage. It divides
the unit cost in shadow prices by the shadow price of output. The indicator is similar to
the Domestic Resource Cost (DRC) ratio, which is well known in the literature. UCs,
however, measures full costs, whereas the DRC ratio, applied at the firm or industry
level, uses value added and thereby ignores the contribution of intermediate inputs to
comparative advantage.
2.2 Improved domestic competitiveness
Under trade liberalization and globalization, industries are normally expected to become
less profitable in the short run because their (shrinking) protective price margin on output
is usually more important than the one on traded intermediate inputs. However, when the
reduction of import restrictions is accompanied by real currency depreciation, the
combined outcome can be the opposite. This is the situation we observe in Indian
manufacturing during the study period. The real depreciation of the exchange rate, which
4 The method is briefly explained in the appendix.
9
eliminated the substantial overvaluation of the 1980s, had the consequence of raising the
price of tradable products, which counteracted the price-reducing effect of cutting the
tariff. The unit cost ratio in terms of domestic prices (UCd) declined from an average of
1.001 in 1987/88 to 0.984 in 1997/98. This means that the total unit cost, including the
opportunity cost of own capital, declined from rough equality with the price to a level,
which is two percent below the ex-factory price.
Table 2: Domestic competitiveness, measured as unit cost ratio in domestic prices (UCd), in Indian manufacturing industries UCd UCd
1987/88 Rank 1997/98 Rank Food industry 20,21 0.9899 9 0.9881 8 Beverages & tobacco 22 0.9279 1 0.8770 1 Cotton textiles 23 1.0454 15 1.0482 17 Wool & silk-based textiles 24 1.0128 11 1.0066 13 Jute & hemp textiles 25 1.0877 17 1.0021 10 Clothing industry 26 0.9620 5 0.9799 7 Wood products 27 0.9932 10 1.0367 16 Paper & printing 28 1.0436 14 1.0359 15 Leather products 29 0.9897 8 0.9749 6 Rubber, plastic, petr.& coal 30 0.9448 2 0.9720 5 Chemicals 31 0.9861 7 1.0030 11 Non-metallic minerals 32 1.0292 12 1.0175 14 Basic metals 33 1.0768 16 0.9510 2 Metal products 34 0.9592 4 1.0056 12 Machinery 35,36 0.9745 6 0.9633 3 Transport equipment 37 1.0334 13 0.9973 9 Other manufactures 38 0.9585 3 0.9675 4 Total Total 1.0011 0.9842 Textile industry 23,24,25 1.0357 1.0304
It can be seen in table 2 that the most profitable industry has been, and still is, the
beverage & tobacco industry. This is due to its high protection, as seen in table 1, a
situation, which prevailed in 1997/98. The least profitable industries at the end of the
study period were Cotton textiles, Wood products and Paper & printing.
The table also shows the dramatic reversal of profitability in the Basic metal
industry, which changed its ranking from the second-least profitable to the second-most
profitable industry.
10
2.3 Changes in export competitiveness
International or export competitiveness is interpreted here as the situation where full unit
costs in terms of domestic prices are inferior to the price on the international market. This
condition is reflected by a unit cost ratio (UCx) inferior to one, as this index divides the
total unit cost in market prices by the border or free trade price. This indicator is shown in
table 3, which also ranks the industries according to cost competitiveness. The fact that
in both years none of the industries comes even close to one, required for
competitiveness, reflects two realities: first, the high cost of protection on tradable inputs,
and second, the use of full opportunity costs of owned capital.
Table 3: Export competitiveness in terms of unit costs and rankings UCx UCx Rank Rank Relative 1987/88 1997/98 1987/88 1997/98 change 20/21 Food products 2.3312 1.3435 4 4
22 Bever.& tobac. 2.3198 1.4018 3 12 D3 23 Cotton textiles 2.3473 1.4794 6 16 24 Wool, silk,etc. 2.5521 1.3472 13 5 R3 25 Jute, hemp etc. 2.6105 1.4530 14 15 26 Garments 2.3177 1.4341 2 13 D2 27 Wood prod. 2.3368 1.5033 5 17 D1 28 Paper, printing 2.7112 1.3949 15 11 29 Leather prod. 2.4248 1.3506 10 7 30 Rubber, plast. 2.2790 1.3422 1 3 31 Chemicals 2.9738 1.3572 17 8 R2 32 Non-met. Min. 2.5279 1.4465 12 14 33 Basic metals 2.9565 1.2600 16 1 R1 34 Metal prod. 2.5146 1.3595 11 9
35/36 Machinery 2.3533 1.2877 7 2 37 Transport equ. 2.3763 1.3890 8 10 38 Other manuf. 2.3870 1.3481 9 6
Total 2.5207 1.3475 Standard dev. 0.215 0.065
11
The indicators of table 3 suggest that for the manufacturing sector as a whole
export competitiveness has increased dramatically, by nearly 50 percent, but unit costs
still exceed free-trade prices by 35 percent, on average. Industries also have become more
uniform in terms of export competitiveness, since the standard deviation has declined
from 21.5 to 6.5 percent.
The “rising stars”, industries which experienced the strongest decline in unit costs
relative to the free-trade price are Basic metals (33), Chemicals (30) and Wool and silk-
based textiles (24). Industries in relative decline, in terms of their ranking, are Wood
products (27), Clothing (26) and the Beverage & tobacco industry (22).
2.4 Comparative advantage
The reduction of industrial protection under the Indian reforms is expected to have at
least three results: (1) First and in the short run it will lower the prices of tradable outputs
and intermediate inputs. Since this effect is normally stronger on the output side than on
the cost side, profitability is immediately reduced. (2) Firms are then expected to react by
cutting costs beyond the cost of intermediate inputs. Such cost cutting usually takes the
form of shedding redundant labour, using the existing capacities more efficiently (to the
extent that sales allow this to happen), and to cut down on non-traded inputs such as
energy, communications and administrative overhead. The resulting cost reduction has a
short-term component reflected by reduced input prices and a long-term component,
which changes the technical coefficients. In the present computations of unit costs we
observe only the short-run effects, however. (3) Firms that do not survive in the market of
increased competition exit or shrink, whereas those succeeding in drastic cost cutting
expand and, in particular, expand into export markets. This changes the structure of the
industrial sector, which is expected to gain comparative advantage. Resources are
expected to flow into those industries with existing or potential comparative advantage
and out of those industries that have neither existing nor a potential for comparative cost
advantage.
12
As table 4 suggests, comparative advantage, as measured by the unit cost ratio in
terms of shadow prices, has been increased over the study period. This means that the
manufacturing industries on average have come closer to comparative advantage by
lowering their unit cost ratio from 1.14 to 1.08, a five percent improvement. The variation
of unit cost ratios also has been reduced from 8.8 percent to 4.6 percent, indicating that
the sector has become slightly more homogeneous in terms of comparative advantage.
This does not exclude cases of industries that have seen their comparative costs rise.
Table 4: Comparative advantage and its change UCs UCs Rank Rank Relative 1987/88 1997/98 1987/88 1997/98 change 20/21 Food products 1.0540 1.0362 2 1
22 Bever.& tobac. 1.1075 1.1135 7 11 23 Cotton textiles 1.1472 1.1666 11 16 24 Wool, silk,etc. 1.1497 1.0675 12 6 R3 25 Jute, hemp etc. 1.3205 1.1628 17 15 26 Garments 1.1008 1.1421 5 13 D3 27 Wood prod. 1.1019 1.1686 6 17 D2 28 Paper, printing 1.2353 1.1322 14 12 29 Leather prod. 1.0800 1.0607 3 5 30 Rubber, plast. 1.0027 1.0982 1 9 D1 31 Chemicals 1.2537 1.0524 15 4 R2 32 Non-met. Min. 1.1987 1.1545 13 14 33 Basic metals 1.3060 1.0511 16 3 R1 34 Metal prod. 1.1114 1.0942 9 8
35/36 Machinery 1.0908 1.0477 4 2 37 Transport equ. 1.1214 1.1113 10 10 38 Other manuf. 1.1100 1.0760 8 7
Total 1.1387 1.0796 Stand.dev. 0.0878 0.0460
The industries that achieved greatest gains in terms of comparative cost advantage
are Basic metals (33), Chemicals (31) and the Wool and silk-based textile industry (24).
They are identified in the table as “rising” (R1 to R3). The industries that have lost out in
terms of comparative costs are Rubber, plastic, petroleum and coal products (30), Wood
products (27) and the Clothing industry (26).
13
The changing structure of the manufacturing sector can be seen in table 5, where
the relative size of each industry is shown in terms of value added and the resource flows
are identified by rising or declining arrows. It can be seen that the expanding industries
have a lower ranking in terms of comparative advantage than the declining industries.
This can be understood as evidence of resources moving towards industries with greater
comparative cost advantage.
Table 5: Value added and its change
VA VA share VA VA share 1987/88 1987/88 1997/98 1997/98 Food products 20,21 25879 0.0903 128105 0.0893 ↓ Bever.& tobac. 22 7360 0.0257 40749 0.0284 ↑ Cotton textiles 23 18741 0.0654 67602 0.0471 ↓ Wool, silk,etc. 24 11121 0.0388 59555 0.0415 ↑ Jute, hemp etc. 25 3051 0.0106 12207 0.0085 ↓ Garments 26 3817 0.0133 32576 0.0227 ↑ Wood prod. 27 1380 0.0048 4241 0.0030 ↓ Paper, printing 28 11306 0.0395 39922 0.0278 ↓ Leather prod. 29 2152 0.0075 12755 0.0089 ↑ Rubber, plast. 30 26206 0.0915 266449 0.1857 ↑↑ Chemicals 31 46472 0.1622 90998 0.0634 ↓↓ Non-met. Min. 32 15148 0.0529 68815 0.0480 ↓ Basic metals 33 33547 0.1171 235113 0.1639 ↑↑ Metal prod. 34 8140 0.0284 35848 0.0250 ↓ Machinery 35,36 46405 0.1620 199728 0.1392 ↓ Transport equ. 37 22059 0.0770 113861 0.0794 ↑ Other manuf. 38 3689 0.0129 25936 0.0181 ↑ Total 286474 1.0000 1434459 1.0000 Textile industry 23,24,25 32912 0.1149 139364 0.0972 ↓
3 Manufacturing exports and their changes
Manufacturing exports have grown substantially during the study period. Annual average
growth amounted to % in nominal terms and % in real terms. This performance is
superior to the growth in the preceding ten-year period. The following table 6 shows the
growth performance of the 15 two-digit-level industries. The number is reduced from the
14
earlier used number of seventeen because in the export statistics cotton, wool, silk and
synthetic textiles, as well as jute & hemp-based products, are aggregated into a single
textile industry (separate from the clothing industry).
3.1 The changing structure of exports
The export performance varies substantially between the industries. Annual average
growth rates during the study period are shown in table 6. The table shows that the Non-
metallic minerals and Basic metals industries experienced the strongest export growth in
terms of growth rates. The most important export industries, however remain the group of
Other manufactures (mainly jewellery), Textiles (Spinning, weaving & finishing), Food
products and the Clothing industry. The growth rates in the second column refer to semi-
real export values. They are given in current US$, such that the Indian inflation
differential is excluded.
Table 6: Export growth in manufacturing industries from 1987/88 to 1997/98 Industry Code Average annual
growth rate (in percent)
Export value 1997/98
Leading products (by increase in export value)
Food 20-21 10.7 3,786,580 Grain mill products Bever.& Tobacco 22 0.7 8,8125 Malt liquors Textiles 23-
24-25 11.0 5,748,024 Spinning, weavg. & finishing
Knitting mills Clothing 26 10.4 3,699,960 Garments Wood prod. 27 9.6 36,002 Furniture Paper & printing 28 15.5 119,950 Pulp & paper Leather prod. 29 3.1 1,284,243 Leather products Rubber, plastic, petrol. & coal
30 3.5 728,176 Tires, tubes
Chemicals 31 15.0 3,546,665 Pharmaceuticals Non-met. minerals 32 20.9 413,051 Div. NMMs Basic metals 33 17.1 1,305,982 Iron & steel products Metal prod. 34 12.7 817,889 Fabricated met. products Machinery 35-36 10.5 1,802,006 Radio, TV, Comm. equipmt. Transport equipment 37 12.7 874,087 Motor vehicles Other manuf. 38 8.1 5,859,121 Jewellery Total 30,109,858
15
3.2 Determinants of export growth
Export expansion may result from a number of factors, some of which are likely a
consequence of the reforms. First, the trade liberalization as reflected by declining rates
of protection, has increased foreign competition and this competitive pressure has forced
the producers to lower their production costs. The reduction in production costs can be
seen as consisting of two parts: secondary input costs decline directly when the protection
of traded inputs is reduced; but primary inputs, in particular labour inputs may also
decline through increased competitive pressure inducing efficiency gains. Firms may also
change their technology through new investments, which is most likely under foreign
investments. The inflow of foreign direct investment may be an important determinant of
export expansion, but the lack of relevant data limits the analysis at this point.
Another source of export expansion may be the reduction of export restrictions,
especially licensing. Unfortunately we do not possess enough quantitative evidence to
examine this potential explanation of export success. Further changes that may have
encouraged exports were institutional changes such as privatisation and hardened budgets
in the case of state-owned enterprises. Here again we miss the quantitative evidence
necessary to analyse this type of reform-related factor.
Based on our approach involving the observation of unit cost ratios we examine
here to what extent export growth is related to the three kinds of competitiveness,
domestic and international competitiveness, as well as comparative advantage. We expect
all three of these indicators to be positively related to export expansion. Domestic
competitiveness, however, should be the least correlated with export expansion, because
it tends to be high under strong protection, which is a disincentive to efforts in favour of
productivity and efficiency gains. Export competitiveness, on the other hand, is expected
to be closely related to export expansion as it measures the profitability in export
markets. However, it is also possible that firms are aware of the potentially ephemeral
nature of price distortions, some of which are implicit subsidies. It will be remembered
that the difference between export competitiveness and comparative advantage is that the
former includes and the latter excludes all kinds of price distortions. Currency under-
valuation, for instance, provides an implicit subsidy to profits, but it may quickly
16
disappear through currency appreciation, and with it the international competitiveness of
whole industries. If firms are aware of these different sources of competitiveness, they
may direct their efforts towards the export market only when they achieve real
competitiveness, which equals comparative advantage.
The following table shows that the most significant determinants of export growth
(dX/X) are productivity growth, measured as the average annual increase in value added
per worker (equation 8) and the change of unit costs at shadow prices (equation 2), which
is our measure of comparative advantage. The coefficient of the latter variable is negative
because unit cost declines imply greater comparative cost advantage.
Table 7: Determinants of export growth Equ. # Dependent var. Independent var. Coefficient t-stats R-square 1 X (1997/98) UCs (1997/98) -1.6 E 07 -1.2 0.11
2 dX/X dUCs/UCs -43.0 -2.6 0.35
3 dX/X UCs (1997/98) 48.7 3.8 0.53
4 dX/X dUCx/UCx -46.5 2.03 0.24
5 dX/X dUCd/UCd -0.18 -0.49 0.018
6 dX/X d(K/L)/(K/L) -0.33 -0.38 0.01
7 dX/X dVA/VA 0.53 1.22 0.10
8 dX/X d(VA/L)/(VA/L) 0.0156 2.89 0.39
While this finding is fully supported by economic theory, it is somewhat
surprising that the export unit cost ratio is a less significant predictor of export growth
than comparative advantage. This may demonstrate that firms are indeed aware of the
potentially volatile nature of export competitiveness as measured by market prices and
including price distortions. Even more surprising is the wrong sign of the regression (#3)
of export expansion on the indicator of comparative advantage itself, rather than its rate
17
of change. Equally, the regression of export expansion (dX/X) on the change in capital
intensity in equation 6 yields a negative sign, which may suggest that export expansion is
not necessarily obtained by investments in more capital-intensive products or processes.
Finally, export expansion is also very weakly correlated with the growth of value added.
In other words, during the study period export growth does not seem to have induced
growth of the industry.
4 Employment growth under the reforms
Employment growth is a crucial aspect of this investigation because of its dual function
with respect to income growth and poverty alleviation, which are, or should be, the final
goals of economic policies. The first function of employment is to generate income and it
amounts to a positive relationship between employment growth and income growth. The
second function is to generate costs, which amounts to a negative relationship between
employment growth and productivity, and hence competitiveness, exports and growth.
Under trade liberalization and globalization the short-run impact is usually employment
reduction, but the long-run effect is expected to be employment growth through raised
productivity and competitiveness. Since we are in this study comparing two points in
time over a ten-year period, where the main policy changes fall into the first half of the
period, we expect to observe more of the longer-run effects. The questions of interest are
then (a) whether there is substantive evidence of employment growth, and (b) whether
employment growth is strongly correlated with export growth.
4.1 Employment vs. productivity growth
As table 8 shows, employment in manufacturing has grown at an average annual rate of
2.24 percent over the study period. The growth rate was particularly high in the Rubber,
plastic, petroleum and coal industry (13.2 percent) as well as in the Clothing industry
(10.5 percent). It was particularly low in Chemicals (negative 4.6 percent), as well as in
18
Cotton textiles (0.3 percent), Non-metallic minerals (0.46 percent), Wood products (0.69
percent), Basic metals (0.77 percent) and Jute & hemp textiles (0.96 percent).
Table 8: Employment growth over the study period
Industry Code
Industry Employed 1987/88
Employed 1997/98
Annual growth
20-21 Food 997,483 1,333,822 0.0291 22 Bev.& Tob. 436,442 599,345 0.0317 23 Cotton text. 834,922 860,690 0.0030 24 Whool & silk 307,606 354,049 0.0141 25 Other textile 196,008 215,986 0.0097 26 Clothing 128,815 369,639 0.1054 27 Wood prod. 70,490 75,502 0.0069 28 Paper & print 290,419 336,664 0.0148 29 Leather prod. 76,389 122,015 0.0468 30 Rubber etc. 209,483 785,571 0.1322 31 Chemicals 549,697 347,792 (0.0458) 32 NMMs 422,720 442,791 0.0046 33 Basic metals 617,278 666,591 0.0077 34 Metal prod. 201,214 278,780 0.0326
35-6 Machinery 810,488 899,492 0.0104 37 Transp. equ. 481,482 551,705 0.0136 38 Other manuf. 77,357 148,383 0.0651
TOTAL 6,708,293 8,388,817 0.0224 23,24,25 Textile ind. 1,338,536 1430725 0.0224
In some of these “low employment growth” industries, however, value added grew
rapidly, thus raising the labour productivity. The greatest increase in labour productivity
is observed in Basic metal (9.6 percent), Wool & silk textiles (6.3 percent), Transport
equipment (6.0 percent) and Non-metallic minerals (5.6 percent).
19
Table 9: Labour productivity and its changes
Industry VA/Empl.
1987/88 VA/Empl. 1997/98
Annual growth*
Annual growth**
20-21 25,944.60 38,772.77 0.0402 0.0251 22 16,863.64 27,447.19 0.0487 0.0217
23 22,446.17 31,708.29 0.0345 n.a.
24 36,152.09 67,906.82 0.0630 n.a.
25 15,564.16 22,816.57 0.0383 n.a.
26 29,630.87 35,577.65 0.0183 0.0182
27 19,578.66 22,678.31 0.0147 (0.0019)
28 38,930.99 47,871.46 0.0207 0.0699
29 28,175.52 42,201.10 0.0404 0.0270
30 125,100.37 136,926.85 0.0090 0.0101
31 84,541.67 105,626.22 0.0223 0.0732
32 35,834.60 62,740.06 0.0560 0.0680
33 54,347.15 142,389.30 0.0963 0.0220
34 40,455.44 51,911.18 0.0249 0.0333
35-6 57,255.51 89,639.73 0.0448 0.0372
37 45,815.00 83,316.08 0.0598 0.0393
38 47,686.70 70,562.47 0.0392 0.0277
TOTAL 42,704.47 69,031.51 0.0480 n.a.
23,24,25 24,588.13 39,323.70 0.0470 0.0314
Notes: Values in Rupees per employee
* based on real values using the GDP deflator
**based on US$ values
The last column in table 9 shows, for comparison, growth rates of labour productivity
based on international statistics using US$ values. They are generally lower because the
Rupee strongly depreciated in real terms during the study period, so that these growth
rates are likely to understate the real growth of labour productivity.
20
It follows from the calculations shown above that during the study period total
employment in manufacturing did not shrink, as it could have been expected under the
competitive pressure, which resulted from the substantive decline of protection and
globalization. Instead, it expanded at a rate of 2.2 percent, with some industries
generating more than or close to 10 percent employment growth. This means that in spite
of the painful adjustments required by trade liberalization and globalization, the
manufacturing sector contributed positively to the growth of income and employment.
5 Conclusion
This paper has attempted to analyse the performance of India’s manufacturing sector
under the reforms of the early 1990s. We found that the level and structure of protection
was drastically changed by the reforms and that this change was accompanied by an
increase of the sector’s comparative cost advantage. Resources seem to have moved in
the direction of industries endowed with comparative advantage. In a number of
industries this has led to increased exports. It is particularly interesting to observe that
export expansion was more strongly driven by comparative advantage than by domestic
profitability, which is high under high protection. Even the profitability on export
markets, but including various distortions on the cost side, was found to matter less for
export success than comparative advantage. Finally, it was also seen that the sector
managed to increase its employment base at an average annual rate larger than two
percent. This suggests that the reforms, although painful for those workers losing their
employment in less successful industries, did not lead to drastic employment loss.
21
References Ahluwalia, M., 2002a, “Economic reforms in India since 1991: Has Gradualism
Worked?,” Journal of Economic perspectives 16(3), 67–88.
Cockburn, J. et al. (1999), “Measuring competitiveness and its sources: The case of
Mali’s manufacturing sector”, Canadian Journal of Development Studies, Fall.
Haragopal, K. Sai, “Trade Reforms and Export Performance”, in Bishwa Nath Singh,
Mohan Pd. Shrivastava, Narendra Prasad, Economic Reforms in India, APH
Publishing Corp., New Delhi, India.
Joshi, V., I.M.D. Little, “India’s Economic Reform 1991-2001”, Clarendon Press,
Oxford, 1996
Panagariya, A.(1999), “The WTO Trade Policy Review of India, 1998”, NCAER, (mimeo).
Panagariya, A. (2002), “India in the 1980s and 1990s: A Triumph of Reforms”, IMF
Working Paper WP/04/43 (mimeo)
Pandey, M. et al. (1998), “Protection of Indian Industry”, NCAER, Delhi, (mimeo).
Rodrik, Dani, 2002, “Institutions, Integration, and Geography: In Search of the Deep
determinants of Economic Growth,” in Dani Rodrik, ed., Modern Economic
Growth: Analytical Country Studies (forthcoming).
Shrivastava, M. P., “A Decade of Globalization, Liberalization and Economic Reforms in
India: An Assessment”, in Bishwa Nath Singh, Mohan Pd. Shrivastava, Narendra
Prasad, Economic Reforms in India, APH Publishing Corp., New Delhi, India.
Siggel, E. (2001), “India’s Trade Policy Reforms and Industry Competitiveness in the
1980s”, The World Economy, vol. 24, No. 2, February, pp. 159-183.
Siggel, E., G. Ssemogerere, (2004), “Uganda’s policy reforms, industry competitiveness
and regional integration: a comparison with Kenya”, Journal of International
Trade and Economic Development, vol. 13, No. 3, pp.325-357.
Singh, P., S. Kaur, “Second Generation Economic Reforms: Some Emerging Issues”, in
P.P. Arya, B.B. Tandon (ed.), Economic Reforms in India: From First to Second
Generation and Beyond, Deep & Deep Publications Pvt. Ltd., New Dehli, India
Srinivasan, T.N. and Suresh D. Tendulkar, 2003, Reintegrating India with the World
Economy, Washington DC: Institute for International Economics.
World Trade Organization (1998), Trade Policy Review: India, Geneva, WTO
Secretariat.
22
Appendix
Indicators of competitiveness and comparative advantage
The indicator of competitiveness used in the present study is a unit cost ratio (UC), defined as total
cost (TC) divided by the value of output (VO), which in turn equals output quantity times the ex-
factory price. For domestic sales, the ex-factory price is the domestic market price (Pd), which is
typically higher than the international price of a similar imported product by a margin equal to the
nominal rate of protection. For export sales, on the other hand, the ex-factory price is equal to the
international (fob) price (Pw).
This particular definition of the unit cost ratio serves a double purpose. First, it helps to
overcome the differences in product mix and quality that make inter-firm comparisons always
problematic. We assume that the output price is usually proportionate to the quality attributes of
products. Therefore, when two firms have the same total cost, but one produces a higher-quality
product with a higher price and total output value, its unit cost ratio would be lower, implying that it
is deemed to be more competitive than the other firm. Second, it makes the unit cost indicator
independent of the data of an international competitor, whose cost we would otherwise need for
comparison. We assume, therefore, that the international price (Pw) corresponds to the unit cost of a
typical international best-practice producer. The fact that Pw is measured as the border price (cif)
means that the benchmark for international comparison includes the transport cost to the border and
therefore a margin of natural (geographic) protection. Our criterion for international competitiveness
or export competitive advantage is then
(1) UCx = TC/(Q Pw) ≤ 1
meaning that a firm is deemed to be competitive if its cost per unit of output is less or equal to the
free-trade price of an equivalent import. This concept of cost competitiveness is multilateral, as
opposed to a bilateral firm-to-firm or country-to-country comparison, but it allows bilateral
comparison as well. For instance, if UCxA > UCxB > 1, then neither country A nor country B, is
export competitive, but B is more competitive than A.
Domestic competitiveness, as defined earlier, means a cost advantage under protection. In
this case the denominator of the unit cost ratio is the output value at domestic prices (VOd = Q Pd),
so that the criterion of domestic competitive advantage becomes
23
(2) UCd = TC/(Q Pd) ≤ 1.
For those firms that export part of their output, Pd of the exported output equals Pw.
In both indicators, UCx and UCd, total cost (TC) includes the interest paid on borrowed
capital as well as the opportunity cost of own capital, taken as the capital stock minus outstanding
debt times the market interest rate. UC exceeds unity if the rate of return is lower than the interest
rate, and it is less than one if the rate of return is higher. Clearly, the indicator sets a high standard of
competitiveness, because the criterion implies that the firm is able to replace its total capital stock by
borrowing at the current interest rate. In times of high interest rates, this may be difficult even for
otherwise truly competitive firms. The indicator has, therefore, this long-run characteristic.
The most important distinction, however, and the hallmark of our method of analysis is the
one between competitiveness and comparative advantage. While competitiveness is understood as a
cost advantage based on market prices including various price distortions, subsidies and penalties,
comparative advantage corresponds to a cost advantage at equilibrium prices. In order to measure
comparative advantage we replace all prices, in output as well as all inputs, by shadow prices. A firm
or industry has then comparative advantage if the unit cost ratio in terms of shadow prices does not
exceed unity:
(3) UCs = TCs/(Q Ps) ≤ 1
where TCs is total cost in shadow prices and Ps is the shadow price of output. For tradable goods, the
shadow price is usually equal to the international price (Pw), but adjusted for any distortion of the
exchange rate. TCs is the sum of all cost components adjusted for all price distortions and subsidies.
It is now evident that the concept of competitiveness differs from the one of comparative
advantage only by including the sum of all price distortions. When UCd is smaller than UCs, the
price distortions act as subsidies; when UCd exceeds UCs they act as penalties. Since price
distortions affect both inputs and outputs, they have the opposite effect on the cost and output sides.
A tariff on output lowers the unit cost ratio (i.e. increases domestic competitiveness), whereas a tariff
on tradable inputs raises it and thereby lowers competitiveness. This shows that in the protected
domestic market a producer is more competitive than under free trade, as production tends to be
more profitable under protection. But comparative advantage, which is the real core of
competitiveness, is not affected by the existing price distortions. However, as a consequence of
protection and other distortions, input coefficients at shadow prices may be affected as well. In other
words, price distortions may lead to lower efficiency and loss of international competitiveness in the
longer run.
Finally, total unit costs net of distortions are broken down into four components, tradable
inputs, non-tradable inputs, labour cost and capital cost, and the distortions are calculated and added
24
to the unit costs at shadow prices to obtain unit costs at market prices. This leads to the following
schema showing how UCd, UCx and UCs related to each other:
(4) VITs/VOs (Shadow unit cost of tradable inputs) +VINs/VOs (Shadow unit cost of non-tradable inputs) +LCs/VOs (Shadow unit cost of labour inputs) +KCs/VOs (Shadow unit cost of capital inputs) __________ = TCs/VOs = UCs (Total unit cost at shadow prices) +dpe (Exchange rate distortion of output) +dpj (Tradable input price distortion) +dpje (Exchange rate distortion of tradable inputs) +dw (Wage rate distortion) +dpk (Capital goods price distortion) +dr (Interest rate distortion) +ds (Direct subsidy, negative) ___________ = TC/VOw = UCx (Total cost per unit of output at international prices) +dpp (Output price distortion) ___________ = TC/VOd = UCd (Total unit cost at domestic prices)
In other words, total unit cost in shadow prices (indicator of comparative advantage), augmented by
all cost distortions, adds up to unit cost per output value at free-trade prices (indicator of export
competitiveness), and adding the output price distortion leads to unit cost in domestic prices
(indicator of domestic competitiveness). This accounting framework serves us to identify, with some
limitations, the sources of competitiveness. The distortions are all expressed as proportions of unit
costs, so that the highest proportions indicate the strongest influence on unit costs.
25