1. Introduction                        

In Vietnam, the textile and garment (T&G) industries have made impressive performance over the past years. The growth rate of the industries has been high at 10 percent per annum as opposed to the country’s annual GDP growth rate of 7.6 percent during the 1990s (GSO, 2001). The T&G industries have played an important role in the economy in general and in the manufacturing sector in particular. The share of textiles and garments in manufacturing output has been around 10.5 percent (GSO, 2001). The share of T&G exports out of the country’s total exports has been around 14.5 percent (GSO, 2001). In 2001, the export turnover of the T&G industries was about USD 2 billion, accounting for 13.3 percent and 16.6. percent of total value of exports and non-oil exports respectively, making the industries one of the leading foreign exchange earners for the country (World Bank, 2002b). In terms of employment generation, the industries take up 25 percent of total industrial employees (Bui, 2001).

Despite significant achievements, there exist a number of constraints to further development of the industries. Although the growth rate of textile and garment industries is high, it has been uneven, with the garment sub-sector growing at 21 percent per annum on average, as opposed to 6.1 percent for textiles (GSO, 2001). According to Hill (2000), although this record is impressive, it is not outstanding in international terms. While the share of garment industry in terms of total industrial output has increased slightly, the share of textiles has declined. This decline may be due to the fact that the textile industry has not kept pace with new trends in demand. Its product is of low quality due to outdated technology. The domestic market for textiles has been dominated by imported goods, especially cheap smuggled textiles from China. The poor performance of the textile industry is widely believed to be caused the obsolescence of technologies (IE, 2001b). The garment industry is also not without problems. Most garments firms operate on low value-added segments of export markets. They often work on a CMT (cut-make-trim) basis where buyers find domestic suppliers, and provide them with product specifications and imported fabrics (Hill, 2000). Consequently, the value added is low. In addition, increasing international competition also poses more downward pressure on prices and costs of textiles and garments firms. The competition has even intensified since the countries in the region devalued their domestic currencies in response to the Asian crisis. All these factors threaten the viability of the textile and garment industries and lower their profitability. In order to rise to these challenges and to compete successfully in the market place to survive and develop, T&G firms must make proper assessment of their competitive position and identify main factors that affect their competitiveness and performance. Such analysis may prove to be useful for policy in terms of making the business and policy environment more favourable for these labour-intensive export oriented manufacturing industries.

Competitiveness is an important topic and therefore is addressed at length in the economic literature. However, there is little agreement about the precise meaning of competitiveness. Moreover, measuring competitiveness is not an easy task. In analysing competitiveness, this study will explicitly adopt a narrowly defined micro approach with a strong emphasis on cost competitiveness. Cost-based competitiveness can be understood in a narrow sense at microeconomic level which is evaluated by the firm’s ability to sell its products profitably in a given market (Cockburn et al., 1998). With this definition of cost competitiveness, we can go ahead with competitiveness analysis for domestic firms without having to identify the firm’s competitors. This is especially relevant for competitiveness analysis of T&G firms where there are so many suppliers in the product market which can therefore reasonably be assumed as perfectly competitive with firms not being able to identify the whole set of their competitors.

In this study, textiles and garments will be treated as two separate industries although in Vietnam they are officially classified as belonging to a single industry. There are two reasons for such a separation. First, the T&G industries invariably use different technologies. The textile industry is characterized as relatively more capital-intensive while the garment industry is featured as more labour-intensive. Second, in terms of trade policy, the textile sub-sector may be considered as a proxy for a range of import competing manufacturing industries while the garment sub-sector can represent labour-intensive export-oriented manufacturing. Analysis and comparison of these industries therefore may provide useful information for policy making in regards of  trade, exchange rate, competition and infrastructure development etc.

Because the study is concerned with cost-based competitiveness, it will focus on (i) surveying theoretical and empirical issues on how to measure firm’s cost competitiveness; (ii) conducting analysis of cost competitiveness to assess the current level of cost competitiveness of T&G firms in Vietnam; (iii) identifying the major determinants of firms' cost competitiveness; and (iv) simulating how proposed policy changes will affect cost competitiveness of T&G firms.

In doing this study, both primary and secondary data are employed. The primary data used for qualitative and quantitative analysis are extracted from the Vietnam Textile and Clothing Competitiveness Survey conducted by the Institute of Economics in late 2001. The survey collected data on 150 T&G firms in the North and South of Vietnam. The secondary data employed for analysis of the performance of T&G industries and the business environment are collected from various sources such as published and unpublished research reports, working papers, articles, and legal documents.

2. Methodology and Data

2.1 The Concept and Measurement of Cost Competitiveness

The concept of competitiveness is widely used by economists, policy-makers, and businessmen. However, there is no consensus as to what is meant by the term. There exist various definitions, approaches and indicators of competitiveness. The concept of competitiveness that is used in this study is the one of firm’s level (microeconomic) cost competitiveness at market prices proposed by Siggel and Cockburn (1995), Cockburn et al. (1998), Siggel (1998), and Siggel et al. Cockburn et al. (1998) provides a simple, but useful definition “competitiveness is the capacity to sell one’s products profitably. For a producer to sell profitably, i.e. to be competitive, his costs per unit of production, also called average cost, must be lower or equal to the market price”. Should any cannot meet this test, it cannot sustain its market position and it would eventually go out of business. Thus, to determine whether a firm is competitive we examine its unit costs relative to the market price. Defining as such, the concept of competitiveness is applicable to analysis of the firm’s competitiveness at both the domestic and international levels. It should be stressed that as this definition of competitiveness relies on market price, competitiveness understood in this sense is a combination of real competitiveness determined by comparative advantage, and various price distortions caused by tariff and non-tariff barriers to trade, subsidies and penalties which Siggel (1998) terms “virtual competitiveness”.

The indicator of competitiveness used in this study is a unit cost ratio (UCR):

UCR = TC /PQ = TC/VO                                                                                (1)

Where: VO is the value of output and VO = PQ.

            P is the ex-factory price; Q is quantities sold.

Defined as such, unit costs are being measured in monetary unit of production. Monetary unit costs will be used in the rest of the study. Thus, we will use the term “unit costs” to denote monetary unit costs and only add the term “monetary” just in case there may be confusion. The applied competitiveness criterion (equation 1) becomes:

UCR £ 1                                                                                                             (2)

Firms are profitable and competitive in the market place if their unit costs are not higher than unity. This measure of cost competitiveness corresponds to the profitability of the firm. As total cost includes the opportunity cost of own capital, taken at the average market interest rate, the firm is deemed to be competitive, i.e. UCR is not greater than unity, if the rate of return exceeds the opportunity cost of capital. Competitiveness in this sense means that the price covers full cost, including the full opportunity cost of capital and is, therefore, a long-run criterion.  If sunk costs are excluded from calculation, UCR>1 means that firms are not competitive even in the short run.

2.2 Firm’s Reactions to Changes in Relative Prices

A policy change normally alters relative prices that the firm faces. In one extreme case, firm does not respond to this change and in the other extreme, firm will give a full reaction to the changes in relative prices. The former may be thought as a very short-run effect of a policy change, when time is too short for the firm to adjust. This interpretation however has limited policy relevance. A more useful interpretation would be that firms are too constrained, in terms of access to resources including capital, labour, technology, information etc., to have any reaction to the changing environment. On the contrary, the full reaction scenario can be thought as the case of market perfection. The reality in economies in transition like in Vietnam should be something in between. For policy, it is useful to compare the two extremes, i.e. not react and fully react, as the difference in firm’s profitability would provide an upper bound for the firm-level cost of market imperfections.

Calculation of the impact on firm’s competitiveness of a policy change is straightforward, but the methodology of taking into account firm’s full reaction needs some detailed. As pointed out in Nicholson (1992), Tybout et al.(1996) and Institute of Economics (IE) (2001c), the task of modeling firm’s reactions to changes in prices of inputs, outputs and other production factors basically amounts to estimating the firm’s production function and/or cost function. For this purpose, the Cobb-Douglas production function is employed. With the assumption that production factors and inputs are combined at two levels, the derivation of the production function is relegated to estimate the Cobb-Douglas value added production function. Then, the impact on firm’s unit cost of changes in relative prices can be measured through  the cost function with an assumption that firms are profit maximizing and facing no constraints in achieving this.    

To analyze the reactions of firm, Tybout et al. (1996) start with the firm’s long run cost function, which allows firm to attain minimum cost (C) at a given output (Q), productivity level (A), and vector of input prices for intermediate goods (I), labour (L) and capital (K) (P_I, P_L, P_K ):

            C = f (Q, P_I, P_L, P_K, A)                                                                                          (3)

Then by Shephard’s lemma, the first derivatives of the cost function are the cost-minimizing input demand functions:

   d lnC = (1/h)d lnQ + s_I (d lnP_I) + s_L(d lnP_L) + s_K(d lnP_K) + (dlnC/dlnA)d lnA           (4)

 where s_I is the share in total cost of the ith factor and h is the elasticity of output with respect to cost, i.e., returns to scale, or also called elasticity of scale. Normalizing by the value of output, they obtain a standard decomposition of the sources of growth in cost per unit revenue:

d lnC – d ln(P_QQ) = (1/h - 1)d lnQ + s_I (d lnP_I - d lnP_Q) + s_L(d lnP_L - d lnP_Q)

     + s_K(d lnP_K - d lnP_Q) +  (dlnC/dlnA) d lnA                                   (5)

Thus, growth in unit costs is equal to the sum of an output growth effect, three relative input price growth effects, and a productivity growth effect. Note that, under constant return to scale (h = 1), the output growth effect disappears because efficiency does not depend upon the scale of production.      

All  terms in equation (5) can be calculated from our dataset. Specifically, the last term of (5) can be derived from (i) the explicit cost function and (ii) estimation of TFP growth rate (i.e. dlnA). In this study, proportionate change in productivity (dlnA = dA/A), is derived from secondary sources (IMPR, 2002b), which in turn made estimates on the basis of the same panel dataset covering 4 years 1997-2000. According to this source, average annual TFP growth rate (dlnA) is 5.2% for the textile sub-sector and 9.1% for the garment sub-sector. The cost function leaves the first term of (5) (namely dlnQ) indeterminable. To model the firm’s reactions to price

changes, one should explicitly derive the supply function. After manipulating, we get:

This elasticity of cost with respect to productivity (TFP) can be calculated based on data in the base year, and parameters estimated in the value added production function.

where  R₀ =[q_va*A*a ^a *b ^b];  g = (a+b);  Q – output (quantities) in the base year; Prices of output, capital, labour, intermediate inputs are all normalised to 1 and therefore dP –output price growth rate; dw₁ – growth rate of price of capital; dw₂  - growth rate of price of labour. Besides, as output price is normalised to 1, output in physical unit as so defined in the base year is equal to the output value.

For the Cobb-Douglas function, g is return to scale and therefore g = h, the first term of (5) - the output growth effect - is equal to:

where  R₀ = [q_va*A*a ^a *b ^b]

2.4 Simulation Scenarios

The simulation exercise aims at quantifying the effects on firm’s competitiveness of changes in policies that are envisaged in Vietnam’s policy reform agenda. Twelve scenarios are framed based on these proposed changes and are summarised in Table 1. The purpose of the simulation exercise is to quantify the effects of on competitiveness of textiles and garments as mentioned above. At the same time, as long as the proposed policy changes aim to remove policy-induced price distortions, this exercise also allows to assess how price distortions caused by existing government policies affect cost competitiveness of T&G firms.

Table 1: Simulation Scenarios and Expected Price Effects

    Source: Scenarios are taken from IE (2001c) with some modifications by the author

It should be noted that firm's reactions (by substituting between labour and capital in creating value added) to changes in relative prices of inputs, output and production factors can take place only under scenarios 1 (changes in tariffs) and 5 (exchange rate realignment), and the two combined scenarios 6 and 7. As long as this study makes a focus on the comparison between “response” and “no response” cases, only the simulation results of these scenarios will be presented and discussed in the next section. The simulation results of the other scenarios (i.e. 2, 3 and 4) will be considered in combination with other policy changes under the two combined scenarios 6 and 7.

2.5 Some Technical Aspects

There are some definitions and technical issues that need to be made explicit. Firstly, in the study, the threshold of 50% clothing out of total production value will be used to separate out textile and garment sub-sectors. In addition, with regard to size of firms, Vietnamese definition is used, according to which a firm with employment of less than 300 workers and capital of VND 5 billion capital or less is considered to be of small or medium scale (small), otherwise is of large scale (non-small). And finally, a firm that has export share higher than 50% of total revenue is defined as outward looking (or export oriented) firm, and inward-looking firm if otherwise. More detailed discussion of these definitions can be found in IE 2001c.

Secondly, the simulation exercise will be done in two steps. In the first step, firm's reactions to changes in relative prices of inputs, output and production factors have not taken into account. As already mentioned, these results are considered as short term or first-round effects only, or of more interest, they can be interpreted as the extreme case of market imperfections that do not allow firms to respond to change. Two indicators of unit cost are calculated. The first indicator includes cost firm’s own capital, i.e., depreciation plus opportunity cost whereas the second indicator does not. Firms with unit cost including cost of firm’s own capital of 100%, or lower (higher) can be regarded as competitive or financially viable (uncompetitive, or non-viable) in the long run as these firms can (cannot) recover all costs. Firms with unit cost exclusive of firm’s own capital higher than 100%, can be seen as uncompetitive (or lacking financial viability) in the short run as these firms cannot recover even variable costs. In the second step, we incorporate firm’s full reactions to price changes into the calculations and make comparison between the two extreme cases.        

The value added production function is estimated separately for textile industry and garment industry with the following results.

For the textile industry :

log(VA) = 0.864 + 0.421^* * log(capital)  + 0.535^* * log(labour)                  (6) 

Adjusted R² = 0.79; Number of Observations = 34

For the garment industry:

log(VA) = 0.474 + 0.137^** * log(capital) + 0.808^* * log(labour)                  (7)

Adjusted R² = 0.67; Number of Observations = 96

Where: VA: value added;  labour: labour input; capital: capital input

                        * , ** denote significant at 1%, 10% level, respectively

Diagnostic tests for omitted variables and heteroskedasticity for (6) and (7) have passed. Statistical tests under (6) and (7) do not reject the null hypothesis that both T&G firms have constant returns to scale. This means that in the long run when firms face no constraints in choosing the mix of production factors, the return to scale is constant (h = 1) for both groups of firms. The first term on the right hand side of (5) disappears and as a result, the first round effect is equal to the final effect. In the short run, when capital of firm is fixed, the returns to scale are decreasing. The return to scale is 0.535 and 0.808 for textiles and garments, respectively.

3. Initial Discussions of Cost Competitiveness of T&G Firms

3.1 Cost Structure of Surveyed Firms

A quick cost analysis may provide important information on how firm’s competitiveness would be affected by various types of policy change. It would also be helpful for interpretation of results of subsequent simulations. It should be noted that in subsequent analyses, monetary unit cost will be expressed in percentage term. Therefore, the competitiveness, or equivalently financial profitability, threshold will be 100% instead of 1. Shares of cost components out of total production value of T&G firms are presented in Table 2. The table shows that for textiles, materials make up the largest share, followed by non-tradable inputs, wages and capital inputs. Taxes and contributions account for only a modest share.

Table 2: Cost Structure of Surveyed Firms (Percentage of Production Value)

Source: Author's calculations based on survey data

These values, however, substantially differ from one another for textiles firms of various types. Our calculations show that there are large differences in share of materials between small and non-small firms (36.3% vs. 57.4%), between state owned enterprises (SOEs) and firms of other forms of ownership. Such differences also exist in shares of non-tradable inputs and wages across firms of various  types: inward-looking and outward-looking, small and non-small, SOEs and foreign invested enterprises (FIEs) and non-state enterprises (Non-SOEs), and firms in the North and in the South. These observed differences suggest that policy shocks such as elimination of import tariffs, more competition in the output market etc. will likely exert differentiated impacts across textiles firms of different types.

Table 2 shows a different picture of cost structure of garments firms compared to that of textiles. For garments, wages constitute the largest cost component which are consistent with the labour-intensive nature of the garment industry. Next come non-tradable inputs and  materials. Capital inputs and taxes and contributions account for relatively tiny shares except for the case of inward-looking firms. Its unusually high share of capital inputs (34.6%) suggests that these firms are inefficient in using capital, resulting in poor performance as indicated by exceptionally high unit cost compared to the average unit cost of garments firms. However, it should be also noted that this result may due to the small sample of inward-looking garment firms, which may not be representative for all inward-looking garment firms. In general, however, there are no large differences between various types of firms with regard to capital inputs. Thus, it is expected that changes in general capital cost will have roughly uniform effects across firms of different types.   

The differences in shares of materials, non-tradable inputs and wages between textiles firms and garments firms are large and statistically significant at 1%. These results imply that there are substantial differences between textiles and garments firms and price shocks to these inputs (for instance, reduction in import tariffs, changes in exchange rates etc.) resulting in changes in the relative prices of tradables over non-tradables etc., would have differentiated impacts on these types  of firms.

3.2 Unit Cost and Firm’s  Competitiveness

As already mentioned earlier, unit cost of 100% is defined as  the threshold that divides all firms to competitive (or financially profitable) and non-competitive (or non-profitable) groups. Distribution of unit cost of T&G firms is illustrated in Figure 1.

Figure 1: Distribution of Unit Cost of Surveyed Firms

      Source: Author's calculations based on survey data

The  figure shows a somewhat brighter picture for textiles than garments firms do. More than half of textiles firms (51.2%) is cost competitive and another 14.6% of firms with unit costs of between 100% and 110% - a range being considered as inconclusive due to measurement errors. The peak of distribution of unit cost of textiles firms is located in the range 90-100%. A relatively large proportion of textiles firms (45.7%) has unit cost in the range 90-110%, implying their vulnerability to a shock and/or high sensitivity to changes in the threshold.

As for garments, only 41.3% of firms are financially profitable and another 11.9% of firms are with unit cost of between 100% and 110% - an inconclusive range. A large share of firms, 37.6%, has unit cost being higher than 120%. The percentage of competitive garments firms seems  not to be very sensitive to changes in the threshold as only about 28.4% of firms are located in the range 90-110%.

Closer look at the data reveals that those firms that have unit cost in excess of 120% are large inward looking textiles and garments firms of various ownership. Most of these firms are located in the South.

Calculations of mean and median values of unit cost show that garments firms surprisingly seem to fare worse than textiles firms regarding both mean and median values, which are  113.3% and 107.8%, respectively, for garments; and 106% and 98.3%, respectively, for textiles. The unit cost of textiles firms is lower than that of garments firms, and the difference is statistically significant at 11%. However, this finding is less striking if we take into account the fact that the textile industry is more protected in the domestic market than the garment industry. The textile sub-sector is characterized as inward-looking whilst the garments sub-sector is considered outward looking. Normally, inward looking industries are more highly protected in the domestic market and enjoy many privileges. Our survey confirms it. The calculated NRPs are not high for garments firms (just 4.1%) but quite high for textiles firms, especially those highly inward looking ones (25.9%). ERPs are especially high for inward-looking textiles firms, estimated at 49.2% by Corden method and 116.6% by Balassa method. The ERPs for garments firms are generally low and the ERPs calculated by Balassa method and net ERPs are negative (-0.9%, -21.4%, and –0.9%, respectively). These suggest that the current trade policy puts the garment industry at disadvantage as compared to the textile industry. Insofar as higher level of protection translates into higher level of profitability, or equivalently, lower unit cost, ceteris paribus, it is easier to explain why textiles firms have lower average unit cost than garments firms. This also points out that textiles firms will be likely to get worse off when Vietnam will have to reduce or eliminate tariffs imposed on myriad articles in pursue of further trade liberalization and realizing its commitments under AFTA agreements in 2006, and WTO agreements that are also forthcoming.

Regarding textiles firms, unit costs differ between firms of various types which generally accords with our initial expectations. Outward looking firms perform better (incurring lower unit cost) than inward looking firms with unit cost of 101.5% and 109.8%, respectively. FIEs have lowest unit cost, followed by non-SOEs, SOEs local and SOEs central with unit costs of 104.7%, 105.1%, 107.3% and 108.3%, respectively. However, it is surprising that North-based firms have lower unit cost than those in the South (101.7% vs. 108.7%). More notably, in terms of size, small firms have much lower unit cost (93.6%) than non-small ones (107.3%). As discussed in Hill (1998), scale economies are significant in the textile industry. The higher unit cost of non-small firms may imply that the textile industry may have not yet reached the point at which economies of scale begin to take effect. Another possible explanation is that domestically produced textiles of larger firms have to fiercely compete with smuggled textiles from China, while products of small textiles have fewer foreign substitutes in the domestic market. It should be noted, however, that all the aforesaid differences are not statistically significant at 10% level.

With regard to garments firms, unit costs are also different across firms of various types. Inward looking firms have much higher unit costs (136.8%) than the outward looking (unit cost of 112.2%) and the difference is significant at 10% level. This purports that Vietnamese garments firms, as long as it can be assumed that the sample firms reasonably represent the industry, cannot operate profitably in their home market. They might be loosing their position in the domestic market to imported garments, especially those products smuggled from China, a pervasive practice that is widely acknowledged in Vietnam. By size, small firms have lower unit cost than non-small firms (112.8% vs. 113.4%) but the difference is not large and not statistically significant. Put differently, size appears not to matter in determining firm’s cost competitiveness. By region, North-based firms have much lower unit cost than South-based firms (105.7% vs. 119.3%) and the difference is statistically significant at 5% level. The result is quite surprising as it is usually perceived that the business environment in the South is more conducive to development of firms. Another astounding empirical result is that SOEs are most competitive (profitable) with unit cost of 104.3-109%, followed by non-SOEs with unit cost of 112%. FIEs have highest unit cost of 123%. The magnitude of difference in unit cost between FIEs and firms of other ownership forms is statistically significant at 10% level. Nevertheless, the higher profitability performance of SOEs vis-à-vis firms of other ownership types should be interpreted with care as (i) the level of accuracy in the accounting systems differs across firms of different forms of ownership (FIEs and private firms are often alleged to inflate their costs, deflate revenues through the so-called price transfer practice and consequently understate their profits), and (ii) SOEs often enjoy more favourable treatments and privileges under the existing business environment in Vietnam. SOEs have better access to quotas to EU, Canada and Norway, which leads to variations in revenues across firms with various ownership forms. The effective borrowing rate of SOEs is lower than that of FIEs and Non-SOEs, thereby lower total costs and unit costs. In addition, it should be noted that by saying a firm is more (less) cost competitive than another firm in this way we mean the former firm has better (less) ability to expand it outputs and relative market share than the latter firm. It does not necessarily mean that the former firm will definitely drive the latter firm out of business. In addition, it also does not necessarily mean that the former firm has unit cost lower than 100%.

By and large, initial analyses show some “puzzling” results regarding the financial profitability of firms of various types. However, care should be taken in interpreting these results as all existing price distortions reflect in this unit cost indicator, thus distorting the true performance picture of these firms. Only when all of these distortions are removed can we make firm conclusion with regard to relative competitiveness of these types of firms.

4. Simulation Results and Firm's Reactions

4.1 Changes in Tariffs on Imported Goods

Trade liberalization alters both value of output and cost of tradable inputs. The net effect of trade liberalization depends on whether the former or the latter effect dominates. First round effect (i.e. without taking into account firm’s response) on imported inputs and outputs of T&G firms of reduction and elimination of tariffs under scenarios 1.1 and 1.2 are presented in Table 3.

All indicators show that with the reduction in import tariffs to 5% or complete elimination (tariffs being reduced to 0%), unit costs of firms, regardless of which measure is used, increase but the magnitude of change varies. The percentage change figures show that unit costs of textiles firms increase more than those of garments firms, for example 10% vs. 2.6% in scenario 1.1. These figures show that tariff reduction, or equivalently, further trade liberalization, hurts cost competitiveness of firms in both industries, particularly textiles firms. However, if both T&G firms are broken down by export orientation, it is found that not all textiles firms are hurt more than garments firms. In fact, it is inward looking garments firms that are hurt the most. Next are inward looking textiles firms. Outward-looking firms of both T&G industries roughly equally adversely affected, but less than the first two types of firms. This finding is rather interesting in the sense that it is a bit contrary to the common perception about textiles firms in Vietnam. These firms are generally thought of as those growing behind high protection barriers. They generally cannot survive in a trade liberalization environment. Yet, these empirical results imply that whether firms are textiles or garments, as long as they are export oriented, they can be prepared to adapt to fiercer competitive pressures in markets exerted by the on-going trade liberalization march.

Table 3: First round Impacts of Tariffs Reduction on Firm’s Cost Competitiveness: Percentage Changes in Unit Costs (%)

Source: Author's calculations based on survey data

Table 4 presents the status of financial viability of firms before and after tariff changes.

Under scenario 1.1, for instance, the number of financially viable textiles firms drops from 21 to 15 (equal to 14% of firms) and from 26 to 18 (equal to 19.5% of firms) in the long-run and short run, respectively. For garments, the number of firms that becomes non-financially viable is smaller, with just 3 (or 2.8% of firms) and 2 (or 1.9% of firms) firms becoming non-financially viable in the long and short run, respectively.

Table 4: Changes in Import Tariffs and Financial Viability of Firms

Source: Author's calculations based on survey data

The first round results show that trade liberalization hurts firms in both sub-sectors. In response to the adverse effects, both T&G firms shrink their outputs in both scenarios (see Table 5), hence reducing growth in unit cost.

Table 5: Effects of Relative Price Changes on Firm's Unit Cost Incorporating Firm’s Reactions under the Scenarios 1.1 and 1.2 (%)

Source: Author's calculations based on survey data

              *: Comparison between returns to capital after and before firm's reactions

Reactions of firms via shrinking outputs are large and almost offset the adverse impacts of increase in inputs prices. The productivity effect is of modest magnitude and helps firms to reduce their unit costs as expected. It can be seen from Table 5 that firm's returns to capital improve quite substantially if firms are able to fully respond to external shocks. Also, calculations from our data show that 2.44% of textiles firms and 0.92% of garments firms become viable from non-viable thanks to the ability to fully respond to the changes in relative prices under  both scenarios.

4.2 Exchange Rate Realignment

Depreciation of exchange rate affects unit cost through two channels. On the output side, it raises the value of output and thereby lowering the unit cost ratio. On the cost side, the opposite effect occurs for tradable inputs, machinery and transport equipment, and electricity tariffs. The net effect of exchange rate depreciation on firm's cost competitiveness is normally positive, as long as the ratio of costs of tradable inputs over revenue is normally smaller than the unity.  

There are alternative ways to estimate the misalignment of exchange rate. Based on the PPP-method, which is possibly the simplest one, we estimate that by 2000, the dong had appreciated in real term and the rate of overvaluation is estimated at 0.2%, 7% and 21.6% depending on which year is taken as the base year. Because one cannot be sure of the most appropriate base year, in the simulation exercise, we use the lower bound of zero and upper bound of 21.6%; and the rate of 7% is being used as an in-between rate to test how sensitive unit cost is to changes in exchange rate.

Table 6: First round Impacts of Exchange Rate Depreciation on Firm’s Cost Competitiveness: Percentage Changes in Unit Cost (%)

 Source: Author's calculations based on survey data

Table 6 shows that in the first round the realignment of exchange rate has moderately (under scenario 5.1) and strongly (under scenario 5.2) favourable effects on firm’s unit cost. The input cost increasing effect of currency depreciation is more than offset by the output cost reducing effect. Therefore, the net effect of currency depreciation is to lower unit cost, thereby improving firm’s cost competitiveness. Unit cost reduction for garments firms is more than that for textiles firms. The reason comes from the fact that per unit tradable inputs cost of textiles firms (55.3%) is more double that of garments firms (26%).

Table 7 shows that under the extreme scenario, i.e., exchange rate is depreciated by 21.6%, five (or 12.2%) textiles firms move up to become viable in the long run, three (or 7.3%) firms move up to become viable in the short run. Garments firms benefit more from depreciation than the textiles. The corresponding figures for garments firms are nineteen (or 17.4%) firms and twenty (or 18.3%) firms in the long run and short run, respectively.

Table 7: Financial Viability of Firms after Exchange Rate Realignment

Source: Author's calculations based on survey data

Table 8 presents the simulation results after firm's reactions are taken into account. As depreciation has produced cost-reducing effects, firms respond positively by expanding their outputs. This effect is relatively strong and more than offset reduction in unit cost caused by depreciation. The productivity effect is as modest as expected. As a result, firm's unit cost increases. However, such increase is not a negative sign as increase in unit cost couple with increase in output does not mean reduction in firm's profit. If the market is assumed to be competitive, the marginal cost curve is upward sloping. Therefore, output increase due to a rise in output price under firm’s profit maximising behaviour will lead to larger profit but also an increase in unit cost. This can be seen clearly by the last row of each scenarios which shows that returns to capital with firm's reactions are substantially higher than returns to capital of firms that do not respond.

Table 8: Effects of Relative Price Changes on Firm's Unit Cost Incorporating Firm’s Reactions under the Scenarios 5.1 and 5.2 (%)

Source: Author's calculations based on survey data

              *: Comparison between returns to capital after and before firm's reactions

So far, the simulation results presented in the previous and this sections have shown that import tariff reduction has largest negative effect and exchange rate depreciation has largest positive effect on firm’s cost competitiveness. However, these changes are conducted separately under the assumption that other things are kept intact. To evaluate the relative importance of these changes and capture the interactive effect between those changes, two combined scenarios incorporating all the aforesaid changes simultaneously are simulated with the results to be discussed in the next section.     

4.3 Combined Scenarios

Table 9 presents the combined effects on firm’s unit cost of simultaneous changes in import tariffs on tradable inputs, machinery and transport equipment, capital costs, electricity tariffs, telecom costs, and exchange rate. Under two combined scenarios, textiles firms are badly hurt with increases in unit cost of 10.3% - 13.3% while garments firms are almost left unaffected under combined scenario 6 and benefit under scenario 7. Textiles firms now have higher unit cost than garments firms. However, if we classify T&G firms by their export orientation as shown in the table, impacts of combined changes vary quite substantially within each sub-sector. Not all textiles firms suffer losses. The unit-cost-increasing effect is very strong for inward-looking textiles whereas the effect is very modest for outward-looking ones. The like also happens as to garments firms. Inward looking garments are severely hurt with unit cost increasing by as high as 16.8% - 28.8% whilst the outward looking benefit with unit cost dropping by 1.5% - 8.5%. In short, in firms in both sub-sectors that are exposed more to the international market can weather policy shocks much better than inward-looking firms. One reason may be that export-oriented firms, even most of these firms export on a CMT basis,  have already had to face fierce competitive pressures in the market. Therefore, in order to survive they must learn to quickly adapt to changing conditions in the business environment as well as external shocks.

Table 9: First round Impacts on Firm’s Cost Competitiveness: Combined Scenarios 6 & 7

Source: Author's calculations based on survey data

Closer look at the data shows that under both combined scenarios, the rank of firm’s cost competitiveness by ownership type remains unchanged for the textile industry, i.e., textiles FIEs are most competitive, followed by Non-SOEs, SOEs local and finally SOEs central. For the garment industry, the rank has changed. Garments SOEs local become most competitive, followed by Non-SOEs, SOEs central, and lastly FIEs. Yet differences in unit cost between T&G firms of  various ownership types are not statistically significant at any conventional level. This implies that ownership type may not have any significant impact on firm’s cost competitiveness. In other words, in T&G industries, performance of firms is not determined by their ownership form. This issue will be left to be discussed further in the next section. 

In what follows, we attempt to assess the relative importance of individual policy changes regarding unit cost. To do this, the combined effects of all policy changes are decomposed into changes in effects by individual changes and the residual between the combined effects and the sum of individual effects is regarded as the interactive effect. Although there are two combined simulation scenarios, the direction of changes of these two scenarios is quite similar. Thus, only the combined and interactive effects of scenario 7 is chosen to be presented in Figure 2.

The figure shows that three proposed policy changes, namely elimination if import tariffs, unifying interest rate and raising electricity tariff exert unit-cost-increasing effects on T&G firms. On the contrary, elimination of tariffs on capital goods, lowering telecom charge and depreciation of exchange rate favourably affect firm’s cost competitiveness. The residual change in policy reform impairs firm’s competitiveness. Among various policy changes, elimination of import tariffs and depreciation of exchange rate have strongest effects of opposite directions on firm’s cost competitiveness. For textiles firms, the unfavourable effects cancel out the favourable effects on firm’s cost competitiveness. As a consequence, cost competitiveness of textiles firms on average is eroded. The reverse is applied for garments firms. Their cost competitiveness improves after simultaneous policy changes.

Figure 2: First round Individual and Combined Effects on Unit Cost of Simultaneous Policy Changes: Combined Scenario 7

Source: Author's Calculations based on Survey Data

As mentioned previously, the quantified effects of policy changes are in fact the upper-bound price distortions component of the current cost competitiveness of firms. They are called “upper-bound distortions” because the policy changes are assumed to take place at the maximum level. In addition, in case of tariff changes we are likely to overestimate the true level of protection as we have not taken into account the effects of duty exemption and smuggling due to the lack of necessary information and the complexity of their estimation. It is widely known that smuggled textiles and garments have flooded into Vietnam’s market for a long time. Most of these goods are sold tariff-free. Consequently, the official tariff rates employed in these calculations may be much higher than the price differentials between imported and domestic products. If all the aforesaid effects of policy changes are considered as current price distortions, their signs will be reversed. If we interpret the results in Figure 2 in this sense, we can show that the current trade policy, interest rate and electricity distortions acts as subsidies to both T&G firms by lowering the average unit cost compared to the distortion-free unit cost. On the other hand, tariffs on capital goods, distortions in telecom cost and exchange rate misalignment act as penalties for firm’s cost competitiveness by increasing firm’s unit cost compared to the unit cost without any distortions. In the foregoing method proposed by Siggel and Cockburn (1995), distortion-free unit cost is called shadow unit cost or equivalently, firm’s comparative advantage. Nevertheless, in our analysis besides the above-mentioned distortions we assume that there are no distortions in other cost components such as wages, building rents etc. Therefore, our calculation of unit cost may not reflect the shadow unit cost. That is the reason why we only call such unit cost “distortion-free”.     

With regard to the viable status of firms, Table 10 shows that under combined scenario 6, as high as eleven (or 26.8%) textiles firms and only 2 (or 1.8%) garments firms move down to become non-viable in the long run. Under the second combined scenario, eight (or 19.5%) textiles firms move down to become non-viable whereas nine (or 8.2%) garments firms move up to become viable in the long run.

Table 10: Financial Viability of Firms: Combined Scenario 6 and 7

Source: Author's calculations based on survey data

The estimates of the effect of relative price changes on firm’s unit cost based on equation (9) are presented in Table 11.

Table 11: Effects of Relative Price Changes on Firm's Unit Cost Incorporating Firm’s Reactions under the two Combined Scenarios (%)

Source: Author's calculations based on survey data

              *: Comparison between returns to capital after and before firm's reactions

The productivity growth effect that is generated by price shocks is very modest compared to other effects in all cases. From the table we can observe that the magnitude of all three input price growth effects is rather similar to the percentage change of unit cost under combined scenarios 6 and 7 in the first round. The simultaneous policy changes exert unfavourable impacts (i.e. leading to positive total input price growth) on cost competitiveness of the inward-looking while have favourable impacts (i.e. leading to negative total input price growth) on more outward-looking T&G firms. In response to unfavourable shocks, inward-looking T&G firms shrink their output as illustrated by the negative growth rates of output. This effect is relatively large to partly offset the increase in unit cost caused by input price growth effect, thereby lowering the growth in unit cost. On the contrary, outward-looking T&G firms expand their output in face of favourable policy shocks. They enjoy more gains from policy shocks via expanding their output as long as firms are assumed to face no constraints in achieving that. For these firms, the output growth effect is the dominant effect. It more than offsets the total input price growth effect, therefore, unit cost of outward-looking firms increases. Nonetheless, from the firm’s perspective, such an increase is a positive, not a negative, thing as such increase in unit cost is associated with increase in firm's profit as already discussed in previous sections. As the marginal cost curve is upward-sloping, output increase due to a rise in output price will lead to larger profit but also an increase in unit cost. The increase in monetary unit cost, or equivalently a decline in profit margin, is more than offset by the increase in firm's output and sales. This is clear from the last row of Table 11, which shows change in returns to capital under each scenario. After reacting to price shocks, firm's returns to capital improve. This purports that firms that have already been profitable become more profitable; and firms that have been loss-making become less loss-making when they have time to react to price shocks. Calculations from our data show that six (or 14.6%) and two (or 6.5%) textiles firms, under scenario 6 and 7 respectively, turn from nonviable to financially viable. Put differently, the number of viable firms in the short run increase when firms have time to react to policy shocks and the goods and factors markets are flexible enough for them to make full response.

In a nutshell, in reality firms react to changes in relative prices of outputs, inputs and production factors by altering their product mix and factor proportions, and also scaling up and down production. As a result, they will suffer less loss caused by unfavourable shocks via  shrinking output or gain more from favourable shocks via expanding output. However, the speed of adjustment of firms depends on many factors such as the level of imperfections and rigidities in the goods and factors markets, the degree of substitutability of various inputs and between production factors (e.g. capital and labour) etc. For inward-looking firms in both T&G industries, when they are aware of coming policy changes in favour of more trade liberalization and less protection and privileges, they must attempt to renovate to be able to survive in a fierce competitive environment. For outward-looking T&G firms, envisaging favourable policy shocks will help them seize the chances to expand and develop. In such a process, the level of flexibility of goods and factors markets has a big role to play. This suggests a promising area for further research.

5. Factor Intensity, Firm’s Characteristics and Cost Competitiveness   

The previous sections focus on discussing the degree of cost competitiveness and how policy changes affect cost competitiveness of T&G. This section concentrates identifying determinants of cost competitiveness which is an important area of policy analysis. The determinants will be analysed in different policy settings, namely in the presence of price distortions as well as when these distortions are removed under various policy reforms. To do this, unit cost will be regressed against firm’s characteristics that may have impacts on firm’s unit cost or equivalently, profitability. A variable represented for labour intensity is also included. The labour share in value added is used as proxy for labour intensity, or equivalently, proxy of how modern the firm’s machinery and equipment is once the sector variable (textiles vs. garments) is controlled. Firm’s utilization rate is also used as an explanatory variable to capture the demand for firm’s products, which to a large extent depends on factors that are external to the firm. Regressions with the same specification are carried out separately for T&G firms. Table 12 presents the regression results. There are three regressions conducted for three policy regimes as presented in the table. The first (1) is the existing policy regime. The second (2) and third (3) policy regimes reflect the expected policy reforms and they are, in fact, the two combined scenarios 6 and 7.

Table 12: Regression Results of Firm’s Characteristics on Unit Cost

Source: Author's Calculations based on Survey Data

   *, **, ***: denote statistically significant at 1%, 5%, and 10% level, respectively.          

With regard to the textile industry, from the table it can be observed that firms with less modern equipment, i.e. higher labour share out of value added once other factors are already controlled for, perform at least as well as more modern firms. The result seems surprising  but the reason may be that firms with modern equipment may not use up their capacity but they still have to incur depreciation costs that are often higher for modern machinery and equipment than obsolete ones. Concerning market orientation characteristic, textiles firms with higher export share, which is a continuous variable, have lower unit cost and this result is robust in all three policy regimes. Be it measured by mean, or median value, unit cost of outward-looking firms is always much lower that of inward-looking firms. The impact of size on firm’s unit cost is not statistically significant at any conventional level. Regarding ownership types of firms, the signs of variables SOE central and SOE local are all positive in three cases, but only statistically significant for textiles local SOEs under the last two policy regimes, implying that in the textile sub-sector, local SOEs are less cost competitive than other type firms. Utilization rate does not appear to have statistically significant impact on firm’s unit cost. Only in the first regime, firm’s specific ERP is included in the regression to capture the impact of trade-related distortions on firm’s unit cost as in the other policy regimes, such distortions are completely removed. The current protective trade policy for the textile industry is really effective since it robustly helps reduce the unit cost of textiles firms, hence improving firm’s cost competitiveness. The regression results further confirm findings in section 3.2 where it is found that textiles firms that are inward-looking and large in the South regardless of their ownership types have highest unit cost. This is evidenced by the sign of variables relating to export orientation, firm's size, ownership type and location.

For garments firms, there are two findings that are statistically and economically significant in all three cases. Namely, firms with higher export share have lower unit cost, thereby more cost competitive; and firms in the North have lower unit cost than those in the South other things being equal. By and large, in the garment industry ownership forms seem not to have any statistically significant impact on firm’s unit cost. However, this result should be taken with a grain of salt. Since FIEs are often alleged to over-report their costs (IE, 2001c), their actual cost level may be lower than reported ones. Therefore, the “puzzle” that garments SOEs are more cost competitive than firms of other types of ownership as found in section 2.2 is not really “puzzling” as the difference is not statistically different. In addition, as pointed out in the previous section on various simulation, under the two combined scenarios when many distortions have been removed, the rank of firms of various ownership forms in terms of cost competitiveness have changed. SOEs central are no longer the best performer in terms of cost competitiveness. This may imply that SOEs would be the main losers of expected policy reforms as specified in this study.          

6. Moving up the Value Chain: A Long - term Strategy

In a recent development strategy for Vietnam’s garment industry, one objective that is considered crucial for the industry is to increase the proportion of exports on FOB basis to move up the value chain. A study by MPDF (2000) argues that producing garments for FOB business creates significantly higher margins as the manufacturers can source the fabrics themselves. To evaluate the relevance of this proposition, unit cost of garments firms under the base case (when unit cost include all price distortions), combined scenario 6, and combined scenario 7 (when unit cost exclude almost all price distortions) is regressed against the share of FOB exports (FOB_share). The regression results are as follows.  

Table 13: Results of Regression of Unit Cost against FOB Share

Source: Author’s Calculations based on Survey Data

Only in the base case, the coefficient of variable FOB share is negative. The negative sign of the coefficient indicates that firms with higher FOB export share have lower unit cost. However, this result is not statistically significant at any conventional level as evidenced by the very low value of t ratio. Under the last two cases, the coefficients of variable FOB share turn to positive, which imply that working on a FOB basis is even less profitable than working on a CMT basis. However these results are not statistically significant at any conventional level, too. In a nutshell, these results purport that there is no evidence that firms working on FOB basis are systematically more profitable than those working on a CMT basis. This finding is in accord with the result found in a recent study on cost competitiveness of T&G firms in Vietnam by IE (2001c) but contrary to the contemporary conventional wisdom in Vietnam on the superiority of working on FOB basis. The reason may be that moving up the value chain, i.e., shifting away from CMT towards FOB exports, requires much additional capital and more specific skills, which are still scarce in Vietnam at this stage of development. However, in practice there have been successful stories of doing FOB business. According to MPDF (2000), these successes are derived from developing strong links with buyers who provide investment, equipment, training and on-site assistance. These firms can purchase materials via local and foreign contracts and create their own designs. In addition, higher position in the value added chain helps firms weather the ups and downs and increasing competition pressure in the international market. This hindsight is especially drawn in the wake of the Asian financial crisis. During the crisis, it was reported that orders were frequently cancelled, especially from Korea and Japan and buyers demanded price cut of up to 20% for subcontract orders (Hill, 1998). The latter also meant that firms doing CMT orders almost could make no profit margins. Firms had to accept such orders as they had no other choice. This necessitates the strategy to move up the value added ladder. However, it should be emphasized that this should be considered as a long-term strategy for the aforesaid reasons. Shifting towards FOB exports can enhance cost competitiveness only if firms can improve their financial position and qualitative factors.

7. Conclusions and Policy Recommendations      

This study is devoted to analyzing cost competitiveness of T&G firms in Vietnam. On average, textiles firms have lower unit cost than garments firms, but inward-looking textiles have much higher unit cost than the average unit cost of garments firms. Outward looking firms are more cost competitive than the inward looking. Statistical tests show that the level of export share is the only firm’s characteristic that affects firm’s cost competitiveness robustly.

The study has found that there exist many price distortions under existing policies, which affect the cost competitiveness of T&G firms in Vietnam. Simulation results indicate that the current trade policy, interest rate and electricity distortions acts as subsidies to both T&G firms by lowering the average unit cost. Tariffs on capital goods, distortions in telecom cost and exchange rate misalignment act as penalties for firm’s cost competitiveness by increasing firm’s unit cost. When these distortions are removed, the impacts are mixed for T&G firms.

The textile industry may be classified as an inward looking industry and it has enjoyed gains from the protection barriers. The ERPs calculated from the survey data are very high for textiles firms but generally low for garments firms. These suggest that under the current trade policy, the garment industry is relatively more disadvantaged than the textile industry. As long as higher level of protection translates into higher level of profitability, textiles firms have lower unit cost than garments firms. This points out that textiles firms will likely be discernibly worse off when Vietnam will have to reduce or eliminate tariffs imposed on myriad articles in pursue of further trade liberalization. 

Trade liberalization in the form of lowering tariff and non-tariff barriers hurts cost competitiveness of firms in both industries. However, the simulation result reveals export oriented firms can weather trade policy shocks better. In addition, under the existing trade regime, outward-looking firms are still more competitive than inward-looking firms. This creates a room for trade policy reforms to step in. If Vietnam is to achieve at least past export success, its trade policy should be more neutral to, if cannot be more favourable for, export oriented industries.

When firm’s reactions to changes in relative prices of inputs, output and other production factors are incorporated in our calculations, the results suggest that in practice firms react to changes in relative prices by altering their product mix and factor proportions, and also scaling up and down production. As a result, they will suffer less loss caused by unfavourable shocks or gain more from favourable shocks. However, the speed of adjustment of firms depends on many factors such as the level of imperfections and rigidities in the goods and factors markets, the degree of substitutability of various inputs and between production factors (e.g. capital and labour) etc. The level of flexibility in goods and factors markets assumes an important role. This is especially important as these reactions of firms play an important role and their effects are relatively large compared to the first round effects.

It has been found that whether price distortions are included or excluded in unit cost indicator of T&G firms, there is no evidence that SOEs, especially garments SOEs, are systematically more or less competitive than firms of other ownership forms. This may purport that in highly competitive markets as T&G markets, ownership type does not matter in determining firm’s cost competitiveness but competition in product market matters. The role of the government is to ensure a level playing field for firms of all types of ownership. In such a highly competitive market, only best performing firms can survive. International competition will discipline all firms, regardless of their ownership form.

In a recent development strategy for Vietnam garment industry, one objective that is considered crucial for the industry is to increase the proportion of exports on FOB basis to move up the vale chain. However, the study has found no evidence that firms working on FOB basis are systematically more profitable than those working on a CMT basis. This finding contrary to the contemporary conventional wisdom in Vietnam on the superiority of working on FOB basis. The reason may be that moving up the value chain, i.e., shifting away from CMT towards FOB exports, requires much additional capital and more specific skills, which is still scarce in Vietnam at this stage of development. Shifting towards FOB exports can enhance cost competitiveness only if firms can improve their financial position and qualitative factors. Moving up the value ladder should therefore be considered as a long term strategy.