In the course of our joint research, we have hosted many workshops, symposiums and policy dialogues. On these occasions, the same policy question always comes up regardless of which industry is being discussed, whether electronics, motorbike, textile, or steel.

Everyone agrees that Vietnam needs to deepen its industrial structure and strengthen its competitiveness. Everyone also agrees that this must be done under the increasing pressure of market competition and international integration. There is also a call for building "supporting industries" that supply parts and intermediate materials to final assemblers. While these goals are easy to endorse, experts and policy makers disagree as to how they should be achieved. One view advocates relatively strong state leadership in harnessing the market on the ground that the private sector alone will not realize the ideal speed or scope of industrialization. In contrast, the alternative view considers the dynamism of the private sector (especially FDI) to be the main engine of growth while the state's role is limited to removing obstacles (usually itself) and realizing the full potential of the market without overpowering it. In the meetings we hosted, Vietnamese participants were often divided between the two views while the Japanese team invariably upheld the second view.

Vertical orientation versus horizontal orientation

Visually, and at the risk of oversimplification, the problem can be re-stated as follows. At present, Vietnam's modern industrial base is small and concentrated in labor-intensive assembly-type processing (gray ovals below). This sector, which is relatively downstream in the production process, is currently dominated by FDI firms and SOEs with foreign connection. Industrial agglomeration is still small even in this sector. Beyond this sector, Vietnam's ability to produce high-quality materials and parts is severely limited. Vietnam also lacks the capacity to design industrial products or market them internationally.

Given this quite empty industrial structure, two alternative strategies are contested.

(i) Vertical orientation. The first strategy is to accelerate the growth of domestic upstream industries (red arrows). The use of domestic inputs is strongly promoted by various policy measures including those that deliberately distort incentives. Localization requirement which penalizes non-compliance is a typical example. If private investment is not forthcoming, the state is urged to invest in desired upstream industries. According to this strategy, the market mechanism should be modified significantly to achieve the national goals. These goals typically include a full-set and "balanced" industrial structure, alleviation of balance-of-payments pressure due to importation of inputs, and national self-sufficiency. It is argued that growth which depends heavily on FDI and market forces will be too slow and biased toward simple manufacturing with low value-added.

(ii) Horizontal orientation. The second strategy contends that Vietnam should refrain from aiming at full-set industrialization in the age of globalization. Instead, this view recommends full utilization of its abundant and highly skilled labor and doing what it already does even better (blue arrow). More specifically, garment, footwear, electronics assembly, food processing, and other light industries should be the driving force of Vietnam's industrialization. Even though exports of these products are rising, they still suffer from low efficiency and remain far short of full potential. If great dexterity is combined with proper management and marketing skills, Vietnam can become the global champion of such industries. The country can grow much faster and create much more jobs by specializing in downstream processing, instead of investing a huge sum in capital-intensive industries. This is because Vietnam's dynamic comparative advantage lies in labor- and skill-intensive industries, and not in capital-intensive upstream industries.

These are the two diametrically opposed views on how a developing country should cope with the forces of international integration.

State-led view versus FDI-led view

Disagreement despite the common goal

Another frequently asked question in our policy discussion is: How can Vietnam build supporting industries if the government does not use rewards and penalties to push localization? Will private and FDI firms automatically invest in such industries? Answering this question requires a string of rather subtle explanations. But the upshot is that it is quite possible to develop supporting industries without forced localization policy.

First, it is not correct to assume that FDI firms are reluctant to buy Vietnamese inputs. On the contrary, all assemblers of the import-substitution type (consumer electronics, motorbike, automobile, etc.) desperately want to procure domestically for their survival. This is because parts occupy the largest share (normally 70-80%) of their production cost, and they can never achieve international competitiveness unless parts cost is reduced. This requires a shift from imported parts which are expensive and slow to arrive to using domestic parts which are cheaper, quicker and more flexible in procurement. Reducing labor cost is not crucial for them since it only accounts for 5-10% of the total cost.

Second, FDI assemblers are actually making efforts to procure domestically. The factory manager of one of the foreign motorbike assemblers says:

The localization ratio of our new model is 59.8%. This includes in-house production as well as purchases from local and foreign-invested parts producers in Vietnam. In an effort to further raise localization, we visit many local factories and rate them according to our supplier evaluation sheet. To receive high scores, they must be extremely conscious of quality and cost. Vietnam has many problems such as absence of materials, restricted imports of second-hand equipment, and high utility charges. We would like to solve these problems in cooperation with the Vietnamese government. On our side, we as a producer are also willing to make more efforts in localizing more complicated parts, gathering information on local suppliers, and transferring know-how and technology. But sometimes the local supplier list provided by the Vietnamese government is imaginary. We go to the address with great effort but the factory does not exist there.

Foreign assemblers feel that they are making much effort to localize, and the criticism of not doing so is very frustrating to them.

Third, there is a perception gap between foreign assemblers and domestic parts producers regarding the quality of domestic parts. Needless to say, foreign assemblers' effort to buy local is conditional; they buy only if local parts satisfy international standards in terms of quality, production management, and punctual delivery. Local parts producers often claim that their products are already good enough or even internationally competitive, while they are not so in the eyes of foreign assemblers. Japanese multinational firms usually set their own quality standards and check the quality of part samples at their laboratories in Japan. No matter how cheap, Japanese assemblers are not permitted to use Vietnamese parts unless they pass the quality test.

Fourth, there are two different business models. On the one hand, Japanese producers consider each motorbike model as an integrated and unique product, which must be designed, produced and marketed separately from other models. Each part is specifically designed for use in just one model and cannot be sold to other producers. On the other hand, Vietnamese producers regard motorbike parts (engine, brake, transmission, fender, etc.) as generic and interchangeable between models. To produce a motorbike, they freely combine different parts available in the market depending on their quality and price. Vietnamese producers cannot understand why a Japanese producer "monopolizes" part producers by banning them from supplying to other assemblers. The existence of these two business models, unique versus generic production, is one of the reasons why a fruitful dialogue between Japanese and Vietnamese producers seems so difficult.

Fifth, parts production is subject to scale economy. In most industries, production of materials and parts in the upstream usually requires higher technology and larger initial investment than final assembly in the downstream. As a result, parts producers will not enter a market unless local demand for their products reaches a certain level. Many Japanese assemblers in Vietnam want parts producers to come to Vietnam, and it is also reported that Japanese parts producers themselves want to invest in Vietnam. But they are hindered by two obstacles: (i) local demand for their products is still too small; and (ii) Vietnamese policy environment is highly uncertain.

It is clear that the goal of localization is common among Vietnamese policy makers, Vietnamese producers, and foreign producers. But they disagree because their perceptions and business models are different. We believe that these differences are surmountable and all parties can work together to accelerate Vietnam's industrialization, rather than continuing to accuse each other. What is required is an opportunity where they can talk to each other frankly and avoid unnecessary misunderstanding (we discovered that Japanese and Vietnamese motorbike firms had never met with each other). Policy makers should sponsor a regular forum for this purpose.

Localization must be based on proper information and cooperation

Vietnamese producers and policy makers think that FDI firms are not making enough effort in localization and technology transfer. FDI firms in turn think that the Vietnamese side is making unreasonable demands without knowing industrial facts or global trends. Both are trapped in an inferior equilibrium. To jump from the vicious circle of mutual distrust to a virtuous circle of constructive engagement, a new approach is needed.

To begin with, both sides should collect and analyze relevant information for designing the localization policy, including:

--Global market trends of the product in question
--Forecasts of global as well as local demand (both final product and parts)
--Assessment of competitiveness of final product in terms of cost and quality
--Current ability of local parts production in terms of cost, quality and capacity
--Possibility of and conditions for attracting FDI parts producers
--Expected capacity and cost reduction of local parts production under concerted effort and support

Based on such analysis, foreign assemblers should indicate their conditional business targets provided that the Vietnamese government will support them and the market will not worsen unexpectedly. These targets should cover production cost and localization (and technology transfer implied by the localization ratio).

The Vietnamese government in turn should offer conditional policy actions provided that foreign producers will make efforts as promised and such policies are consistent with external trade commitments (AFTA, WTO, USBTA, etc.). They should include tariff rates and special measures to encourage domestic parts production (see the next section for more details).

The targets and the policies should be mutually dependent and based on best available forecasts of the industry and the market. The two must mesh closely for successful localization.

Producers sometimes overstate the difficulties they face to entice more favorable conditions, so it is not always necessary to grant all their wishes. On the other hand, if policy becomes too unrealistic, business cannot survive. To strike a proper balance between targets and support, intense dialogue with producers and deep knowledge of the industry are absolutely necessary.

It is important to recognize that market conditions also significantly affect the outcome of localization policy. FDI firms should not be held responsible for shocks that are beyond their control (such as war, recession, natural disaster, unreasonable policy, etc.) In particular, FDI firms should not be criticized for deviating from the "business plan" in their original investment permits. In a market economy, such plans are merely indicative and never a commitment. Business strategy must be constantly adjusted as markets and policies change.

The general menu for promotion measures

Import protection, localization requirement, and tax and financial privileges for domestic production can be used, but only for a very short time. These measures are generally prohibited under AFTA or WTO, and waivers and exceptions are hard to negotiate.

However, without violating AFTA or WTO, the Vietnamese government can do much to improve the business conditions in general and the localization policy in particular. We recommend the following actions:

If these measures are implemented, Vietnam will surely increase FDI receipt and become the factory of the world in labor- and skill-intensive manufacturing. When that happens, and as a result of such agglomeration of assemblers, supporting industries will grow naturally in response to rising demand, supported additionally by a favorable policy climate. This is demand-led localization that we advocate, which must be clearly distinguished from the forced localization policy currently in place.