The Malaysian Model and its Relevance for South Africa

Malaysia has had a strong allure for South Africa and other countries in the region. Of particular interest have been its policies aimed at redressing racially-based economic imbalances and its willingness to question the merits of conventional macroeconomic and industrial policy advice from international financial institutions.

Like South Africa, Malaysia has been ruled since independence by a single coalition dominated by the interests of the majority ethnic group. An election this Sunday will provide a real possibility of political change. The occasion calls for a realistic assessment of what has come to be regarded as the "Malaysian model."

My own view is that a) the divergence of Malaysia's macroeconomic policies from conventional "best practice" has been greatly overstated, b) whatever the merits of "bumiputra" affirmative action policies, the costs have also been large and by now certainly outweigh the benefits, and c) key features of targeted industrial policies have impeded rather than promoted the country's economic growth.

These views are echoed concisely in a Financial Times opinion piece. According to the columnist, David Pilling, failures of past policies make political change a real possibility.

"There is huge anger at entrenched corruption and buddy-buddy crony capitalism. Many Malaysians have come to the conclusion, almost certainly correct, that affirmative action has outlived its purpose and is holding the country back.

"There is increasing recognition, too, that economic performance is not all it might have been....Manufacturing exports have been stalled for years and attempts to build an indigenous steel and car industry have flopped. It has been easier for cosseted businessmen to jostle for lucrative state contracts than to compete internationally."

Pilling's view is that there is a "need to roll back a system based on race and patronage in favour of one based on competition and merit."

Whether political change happens, and whether it will result in meaningful policy reform remain to be seen. But discussions in South Africa and elsewhere need to be informed by a realistic view of the nature and impacts of the "Malaysian model." Otherwise the dti and others will continue to copy Malaysia's failures rather than learning the real reasons behind its successes.

SA's Motor Industry—Once Again

The release of the dti's latest Industrial Policy Action Plan (IPAP) has put the motor industry in the spotlight once again. The dti continues to claim the industry as its greatest industrial policy success. However a recent news report draws attention to IPAP data showing that this sector alone accounts for about 40 percent of the country's trade deficit, and asks what this means for the success of the dti's policies. If government support has created an internationally competitive motor industry, as claimed, why does it show such a large trade deficit?

The implicit answer, supported by a number of online reader comments, is that the industry is not internationally competitive. This is consistent with my own analysis over the years—the ability of MIDP and other government policies to attract investment and promote exports reflects, not the industry's competitiveness, but rather the value of government incentives. 

The MIDP was recently replaced by the APDP. This tweaked some of the details of public support for the industry—production subsidies are now given to domestic sales as well as exports; the dti can now hand out cash investment incentives on a more discretionary basis. But the magnitude of public support has not diminished; and it is determined largely in consultation with the major firms who tell the government how much support they require to overcome the high costs of investing and producing in SA. In other words, the type and magnitude of public support depend on how uncompetitive these firms are in SA.

The danger now is that the dti will seek to solve this sectoral trade deficit "problem" in ways that will  increase the burdens on consumers and taxpayers, and further diminish SA's manufacturing competitiveness. A simple "solution," for instance, would be to impose local content requirements on the industry and restrict imports through higher tariffs or other measures—in other words, return SA to the bizarre and highly costly policies of the Apartheid era.

The government could, and almost certainly will, employ more subtle subsidies and incentives to serve its motor industry clients. A tender document just released by the DPE, for instance, solicits advice on how Transnet and Eskom can be used to increase motor industry profitability. Reducing Transnet and Eskom costs would certainly be good for the entire economy. But to give special privileges to the motor industry (which already benefits from special Transnet pricing deals on its exports) would perpetuate the failures of the current policies that seek to improve competitiveness by subsidizing uncompetitive firms and industries.

Subsidies, Bailouts and Markets

I read with relish the report of Leon Louw's recent blast at the never-ending bailouts of South African Airways (SAA). Mr. Louw, Executive Director of South Africa's Free Market Foundation (FMF), demonstrates how these subsidies distort competition, raise costs and prices, divert resources from more efficient and growth-promoting activities, and are especially burdensome to the poor, from whom the bailouts divert public support. The FMF performs a great public service through well-reasoned commentaries of this sort.

Of course, it is not just state companies and agencies that benefit from public subsidies. Direct government support and regulation are viewed by some as necessary for South African industrial development. The consequences of this approach can be similar to the effects of bailouts of state enterprises.

In this regard, Mr. Louw's remarks reminded me of a brilliant commentary by another FMF member, on the impacts of industrial policy in South Africa's motor industry. In a prize-winning letter to Business Day in 2005 Jim Harris explained how the MIDP (Motor Industry Development Program, recently rechristened as APDP) supports a few companies, but punishes consumers and retards the country's long-term economic development. His remarks were controversial and attracted fierce reactions from some fellow FMF members—at least those associated with the motor industry.

Mr. Harris' letter signalled the start of a vigorous public debate about MIDP that, if nothing else, clarified the nature and magnitude of the subsidies enjoyed by the industry. As with the bailouts of SAA, the MIDP and related subsidies protect a few heavily dependent companies at the expense of consumers, jobs and the poor.

I have been unable to find any follow-up on the FMF website to Jim Harris' early foray into South Africa's most important industrial policy. It would be useful for the FMF to complement the comments on bailouts to state enterprises (and on the US bailout of General Motors) with a discussion of the massive and continuing subsidies to firms in the South African motor industry and other selected sectors.

Making a Mockery of Competitiveness

One of the few genuine innovations in South Africa's APDP, the motor industry support program that will replace the MIDP, is a requirement of a minimum scale of operations in order to qualify for the program's very generous financial benefits. Although the details of the program still have not been gazetted, we understand that in order to qualify vehicle assemblers will have to produce a minimum of 50,000 vehicles per year. The idea is to encourage international competitiveness by achieving some economies of scale in production.

The scheme does raise some questions.

1. Why is it necessary for the government to instruct global auto firms on efficient scales of production? (It is not; but under MIDP and APDP it is not necessary to be competitive in order to make big profits by producing in South Africa.)

2. Will production of 50,000 vehicles per year provide sufficient scale to compete with true international major league plants? (Generally not; competitive plants elsewhere produce at least 100,000 per year.)

3. And of course one might wonder how firms will try to get around the new requirement in order to continue to take advantage of the large subsidies made available by the government (and paid for through the "generosity" of South African vehicle consumers and taxpayers) without ramping up production.

But at least the new requirement appears to be a feint in the direction of encouraging some competitiveness on the part of the local industry.

An announcement by the government-operated East London Industrial Development Zone (ELIDZ) provides an answer to the third question. ELIDZ has told the press that it will soon announce the signing of several tenants in a new "multi-OEM assembly plant" it will build in the zone. What is the purpose of such a plant? To enable companies sharing the facility to produce at volumes of 10,000 vehicles per year, or even as low as just a few thousand and still qualify for the APDP benefits that are contingent on meeting the 50,000 per year volume target. It apparently will "work" as long as the total of all vehicles produced by all tenants reaches 50,000 per year.

How will this encourage increased competitiveness, the stated goal of the new policy? It will not. In fact it will simply replicate the highly inefficient pattern of small scale production of multiple models and other automotive products under one roof that was encouraged by protection in pre-MIDP days and by continued protection and MIDP benefits since then.

More Protection for the SA Clothing Industry?

Following a request from SACTWU, the Clothing and Textile Workers Union, South Africa’s International Trade Administration Commission (ITAC) has reviewed the import duties on garments. Its report concludes that the duty should be increased from 40 to 45 percent, the maximum allowed under SA’s WTO commitments. For several items the current duty is only 20 percent, and yet they would still get the new 45 percent rate. The recommendation covers virtually all clothing used by South Africans from socks and underwear to shirts, jackets, suits, pants and skirts; for men, women, girls, boys and babies.

This follows another recent recommendation and government agreement to reduce the duties on a wide range of imported fabrics from 22 percent to zero.

ITAC argues that these measures will “provide significant encouragement and support to the industry while its extensive efforts to restructure to become more competitive are continuing;” and they will have little inflationary impact.

ITAC does not show its estimates of the amount of support this will provide. My own calculations show that before any of the recent recommendations, i.e. with tariff rates of 40 percent on clothing and 22 percent on most fabrics, SA garment producers were being subsidized at a rate of about 94 percent of value added. In non-technical terms, this means that the tariff structure allowed them to produce garments at a cost 94 percent higher than foreign competitors and still compete in the local market. This is a very high level of support, higher than in just about any other industry in South Africa.

Under the proposed new measures, the subsidy will increase to 114 percent for garment producers that are still subject to the 22 percent fabric tariff. And for those that are able to take advantage of the new rebate that lowers the fabric tariff to zero, the subsidy will be 180 percent. They will be able to produce at almost three times the cost of foreign competitors and still be able to “compete” in the domestic market.

Does this encourage restructuring and increased competitiveness? No, it subsidizes high cost producers, and enables them to survive despite much higher costs than international competitors, with no need to adjust or to improve their competitiveness.

On top of this, the government is mulling even further subsidies to the industry. No details have been made available, but under proposals that have been mentioned publicly I estimate that the rate of subsidy could be over 300 percent.

It is difficult to imagine how firms that require this kind of support could be transformed into internationally competitive producers.

What, then, is the real goal of government support? It seems to be to protect vulnerable and/or a privileged subset of the country’s workers. But is subsidizing continued production and further investments in training and equipment to sustain existing jobs in non-competitive firms or industries the best way to do this? The cost is high. Clothing and footwear account for 5 percent of consumer expenditures in South Africa. For lower income households the percentage is much higher. A 40 or 45 percent tax on such a basic consumer good is a cruel and regressive tax indeed. While an increase in the tariff might provide further short-term relief for garment workers, it will be at the much greater expense of all workers and consumers in the country, and especially the poor.

A government official recently told me that one of the main problems with the SA garment industry is that the equipment is very old. The unwillingness of firms to invest in upgrading this equipment, despite current high levels of protection, would be a rational response to their view that much of the industry will never be able to compete.

I asked about the age of workers in these factories. The answer: “about the same as the equipment.” This might also reflect a rational response by young workers who see no future in this industry. It also reduces the costs of any necessary adjustment if industrial support is gradually withdrawn. An increase in support might not only impose higher immediate costs on taxpayers and consumers, but might draw new workers into the industry and increase future adjustment costs.

What is the bottom line? The government needs to move out of the box that views subsidies to non-competitive producers and industries as the best way to help the country’s workers, taxpayers and consumers. Let's deal directly with what is essentially a social welfare problem – assisting workers in declining firms – and free the government to think more creatively about an industrial strategy that points to the future rather than the past.

Health for Empowerment and Growth

Promoting empowerment of disadvantaged groups through incentives to broaden ownership of capital has had mixed results in terms of social justice and long-term growth. Experience in Malaysia and South Africa for instance shows that such schemes have certainly enhanced the wealth of a select few of the “previously disadvantaged.” But broader trickle-down effects to the poor have been much harder to find. And the costs have been high, with serious questions about the extent to which they might have impeded longer-term productivity growth as a result of implicit taxes on wealth and creation of a rent-seeking business society. As a result, there is growing pressure in Malaysia to abandon the “bumiputra” policy. Some of the loudest voices belong to those who were supposed to be the beneficiaries of the program.

These countries should take note of some new research findings from the US. A recent report by Chay, Guryan and Mazumder shows the startling and large effects of improvements in basic health care for black infants in the 1960s arising from hospital integration. They show that better infant health care explains substantial improvements in educational achievement, independently of other causes such as school integration and changes in family backgrounds.

The lesson for countries like Malaysia and South Africa is clear. If you want to promote empowerment of disadvantaged groups, use basic health and education to tackle the root causes of inequality and the real barriers to economic opportunity. The resulting improvements in human capabilities will pay additional dividends in the form of higher long-term productivity growth as well.

A "New" Program of Auto Industry Support

It's been a long time coming. But in early September the dti finally announced the general architecture of its revised MIDP, now called the Automotive Production and Development Program (APDP). The official announcement is sketchy. Public support of the industry will be extended until at least 2020; the program will be "production neutral" (i.e. it will subsidize both exports and production for the local market) and so will be more WTO-compatible.

Other than that, the details, magnitude and impacts of the program will be very difficult for most observers to fathom. Two major studies commissioned during the three-year review process have not been released for Parliamentary or public scrutiny. The Industrial Development Corporation (IDC) conducted a "cost-benefit analysis" of the proposals for the dti, but that is also being kept under wraps.

Of course there will be few mysteries or surprises for the motor industry. The program was designed in close consultation with them by two teams of dti consultants and by the dti itself. This is what the dti calls "stakeholder consultations." But what about other important stakeholders -- Parliament and the consumers and taxpayers they represent, and who ultimately pay for these and other industry support programs?

Broader stakeholder discussion and understanding require much wider dissemination of information and analysis. Justin Barnes and Anthony Black, who have been involved with the MIDP since its beginning, shed some light on the thinking behind the new program in an opinion piece. 

This was the cue for a short article of my own, in which I draw out some implications of their observations and fill in a few blanks with my own estimates of the costs and magnitude of the subsidies provided under both the old and the new programs. The costs have been and almost certainly will continue to be high; catalytic converters -- canned platinum -- have received as much export support as automobiles; despite an apparent decrease in some of the rates, overall rates of subsidy for both exports and locally sold production will increase rather than decrease under the new program; the new incentives are likely to create a barrier to entry for new producers.

At least some of these consequences are unintended. But whether intended or unintended, appreciated or not appreciated, they are certainly not what has been advertised by industry advocates. 

When policy design and assessments are done behind closed doors, in consultation primarily with only the direct beneficiaries of the programs, questions are bound to be raised. From the information available in this case it appears either that the industry is not competitive and is unlikely to become so, at least for a very long time; and/or that consumers and taxpayers are being taken for a ride by giving unnecessary subsidies. In either case the argument for continued support appears to rest on very weak foundations.

A box in a recent OECD report on South Africa singles out the dti's support of the motor industry to illustrate the dangers of the "capture" of policy makers in the design and implementation of industry-specific industrial policy programs.

This is not to deny the role or importance of industrial policy. But industrial policy needs to be based not just on the influence of powerful industry lobby groups or on the inclinations or dreams of policy makers; it needs to be based on sound economic analysis. To avoid the dangers of capture, it also requires institutions and processes that are transparent and accountable. Policy reviews need to be made public; alternatives need to be presented and properly assessed in terms of their overall economic impacts. And industrial policy needs to be based on a recognition of the capabilities of the policy-making authorities to do the ongoing monitoring and assessment that are represented as a key part of the dti's own Industrial Policy Action Plan.

Unsafe at any Speed: MIDP and SACU

In three articles headed the wheels may come off South Africa's Mail and Guardian draws on a recent paper by Matthew Stern and myself to explain problems with customs revenue sharing in the Southern Africa Customs Union (SACU).

The articles cleverly explore the links between SACU's new revenue sharing formula, South Africa's Motor Industry Development Program (MIDP) and the recent boom in car sales in South Africa, to illustrate how the new formula turns all previous interests in SACU trade policy on their heads. By fattening the government budgets of the four small and less developed SACU members (Botswana, Lesotho, Namibia and Swaziland—the "BLNS countries" in SACU-speak) the formula has turned these countries into avid supporters of a program designed to subsidize South African motor industry producers at the expense of South African and BLNS consumers. The cost of the program to BLNS consumers is now dwarfed by the (far more than) offsetting revenue transfers to BLNS governments (to the tune of 23 percent of GDP in the case of Lesotho and 12 percent in Swaziland). And now poor South African consumers/taxpayers are saddled not only with high priced motor vehicles but also with the burden of huge fiscal transfers to the BLNS.

South Africa's National Treasury, quite naturally, is bemoaning not only the size of the revenue transfers, but also the difficulties they will create in reaching agreement on badly needed rationalization of the SACU tariff structure. Under the new SACU arrangement, tariff rate decisions are meant to be decided collectively by all members. Having been turned into the greatest supporters of South Africa's protectionist trade policies, the BLNS countries are unlikely to favour tariff reductions that threaten their fiscal windfalls. Before complaining too much, however, South Africa should look at its own central role in promoting agreement on this new revenue formula.

Among the cutest parts of the M&G articles is an illustrative calculation of the amount of revenues transferred to the BLNS whenever a South African buys a new imported car. Based on an import duty rate of 25 percent and an assumed transfer of about half of these revenues to the BLNS, the article shows that the purchaser of a new "Volvaru" that cost a dealer R200 000 would pay just over R25 000 to the BLNS, of which about R5 000 would go to the government of Swaziland and R4 100 to Lesotho.

In fact the numbers are much worse than that. The import duty on cars is now 30 percent, not 25 percent. And, according to our estimates of last year, the share of duty collections going to the BLNS is about 88 percent, not 50 percent. This means that a South African buyer of the above-mentioned "Volvaru" would contribute about R52 000 to the BLNS, of which over R11 000 would go to Swaziland and R7 400 to Lesotho (and greater amounts to the governments of Botswana and Namibia).

This tale leaves South African consumers and taxpayers with much to ponder. The consequences of the MIDP and of the new SACU arrangement are surely much different than whatever might have been expected by those who designed them; and their combined impact is far worse than either of them individually. Two bad policies do not make a good one.

South Africa's MIDP

South Africa's Motor Industry Development Program (MIDP) has been one of my favorite topics recently. The Department of Trade and Industry (DTI) and the motor industry have been remarkably quiet as the country awaits the release of the MIDP Review (due sometime last year) and the government's future policy framework for this industry.

However, an industry representative was reported yesterday as saying that government support of the industry is "below par" relative to other countries and will have to be improved in order to make the industry competitive.

How much support has the MIDP given the industry? Answers to this basic question are difficult to obtain. I have estimated that typical subsidies to motor vehicle and/or components manufacturers range from 260 to over 650 percent of the size of MIDP-supported investments.

Support to the motor industry from export subsidies and import duties has almost certainly exceeded R100 billion over the life of the program so far. The cost to consumers has been similar. How can an industry that relies on such levels of support expect to become "competitive"? Surely the solution is not to give even more subsidies.

Australia, after whose program the MIDP was initially modelled, is entering the final phase of its "Automotive Competitivness and Investment Scheme." This program provides modest production and investment-linked subsidies and a phase down of its import tariff from 10 percent at the moment to 5 percent in 2010. The South African tariff is still 30 percent—a long way to go before encouraging competitiveness.

For more on this, including a link to the news story in question, see the Features page of my website.