THE INDUSTRIALISATION OF ZIMBABWE-PAST, PRESENT AND FUTURE

In this paper I look at Zimbabwe's moderately successful industrialisation experience, past, present and future. The lessons that can be drawn from this experience in comparison with what has happened in other countries, both more and less succesful, are: that there is a need for an intelligent state role; that both import substitution and export substitution are necessary; that emphasis on industry need not and must not be mean neglect of agriculture; and that the key problem is how to avoid enclave industrialisation which services urban elites but neglects the rest of the country and the region.

2. the early attainment of a measure of domestic control of the economy.
In contrast to the situation in Zambia with its copper, or Ghana with its cocoa, or a number of other economies in which one primary product was of dominating importance, Zimbabwe did not have a 'real business' that was manifestly the most profitable activity and from which manufacturing industry might have been seen as diverting investment. On the other hand, mining and agriculture were successful enough to make substantial and varied demands on industry for their inputs, from mining équipement for a dozen different sets of geological conditions to irrigation and tobacco-curing equipment, fertilisers and insecticides.
Since Arrighi's seminal essay of 1967(1) the attainment of the white settlers of political power in 1923 has been seen as a crucial event. It is not clear that a desire for more domestic control over investment was a conscious aim of the colonialists at that stage, but political power immediately raised the possibility of investment for the long-term future of the colony, rather than the short to medium-run profitability for metropolitan shareholders, as tended to remain the criterion in other colonies.
Thus began a state-led process of industrialisation favouring settler interests, in part at the expense of the economically dominant British South Africa Company wich still controlled mineral rights and owned the Wankie Colliery and the railway. Local capital in the 1920s and 1930s was predominantly agricultural, but small-scale miners (up to a thousand in number) were competing with the larger foreign-owned mines, and manifacturing industry was beginning to supply a range of consumer goods to the settler population, and to seek protection against imports. The 1930s saw the Tobacco Marketing Act (1936), designed to strenghten the power of tobacco farmers against the monopolistic United Tobacco Company, the establishment of the Electricity Supply Commission, and the setting up of the roasting plant, as a disguised subsidy to small-scale domestic goldminers. In the 1940s major state investments were made in the Rhodesian Iron and Steel Company (now Zisco) and in cotton ginnery. Both of these provided a big stimulus to downstream private manufacturers (as well as to cotton growners in the latter case), and manufacturing industry began a steady rise from about 10 % of GDP before the war to 20 % in 1965, whilst agriculture declined in relative terms to about 15 % and mining to about 7 %. Isolation and the economic conditions of World War II and its aftermath provided natural protection, but during the years of the Federation of the Rhodesias andNyasaland (1954-1963), a complicated tariff structure was introduced, and after gaining preferences in their markets, Southern Rhodesian industry was stimulated by demand from its two less industrialised partners. A complex of opportunities and state responses not available to ordinary colonies thus allowed the development of an industrial structure with few if any parallels outside the independent dominions. However these opportunities were never pursued unambiguously. As industrial capital grew it developed a need for a larger market, and as it was not yet internationally competitive this meant the internal (or federal) market. This in turn required the expansion of black purchasing power through restoring the viability of black agriculture, and thus it came into conflict with the politically more powerful agricultural capital, the white middle class and white workers, who therefore combined to close off that avenue of expansion by the election of the Rhodesian Front in 1962.
Nevertheless, enforced protection during the UDI period gave an unexpected further stimulus to import substitution, and manufacturing's shares of GDP reached 25 % by 1974, and with continuing protection after independence, over 30 % in 1986. Since 1966, tariffs have been relatively unimportant, with quantitive restrictions on imports being imposed so as to bring the total value of imports in line with export earnings. Because of the shortage of foreign exchange, initially arising because of sanctions on Rhodesian exports, any manufacturer who could demonstrate a capability for local production of any item was very likely to gain protection. Therefore it is important to appreciate that circumstances played a major role in creating a climate conducive to the development of industry: although the state was somewhat interventionist as just discussed, 'natural protection', World War II and UDI also played a major role. Indeed, the professed ideology throughout was of a free market, open-economy nature which was at odds with the interventionist, semi-autarchic policies actually being pursued. Thus it is ironic that successful industrialisation followed from a policy that the dominant ideology would have prevented had circumstances not over-ruled. Government however did not play a very large direct role in industry, although its catalytic effect in steel and cotton were probably crucial. Iron and steel were denationalised in the late 1950s, but government again increased its stake to a majority position in the UDI period, after a major expansion resulted in high financial liabilities when the sanctions-busting of the Austrian partner was exposed. The main direct intervention was through the Industrial Development Corporation (IDC) which was set up in 1963 with capital of 1 million from government and private institutions, and charged with aiding development through strategic investments. These were, however, invariably minority investments, and the IDC appears to have acted, without government interference, as a minor investment company.

Background
After independence, industrialisation was identified as the key long-term requirement for economic development. Aspects involved in formulating an industrial strategy included the dominance by foreign and settler capital, the orientation towards production for an elite, the consequent need for an expanded state role, the balance between import substitution and export orientation, the orientation to agriculture and the mining, the pressure exerted by the World Bank, and the need to raise the proportion of the population in employment. I do not have space to discuss all these questions explicitly, and so will concentrate on the role of the state, the impact of the World Bank, and export strategy. This will enable an attempt to be made to assess the character, strengths and weaknesses of the 'Zimbabwe model' of industrialisation. First, however, some factual background.
In the years after 1980 the state slowly developed a somewhat greater role in industry, with the IDC both taking over a number of ailing companies, and becoming involved in a wider range of investments, sometimes of a majority nature; in 1983 it had total assets of about ZW$ 40 million, under 1 % of total industrial assets. In 1985 it was expanded and became 100 % government owned, but remained self-financing, and in early 1988 its capital was increased tenfold to ZW$ 100 million prior to a new expanded role. It now owns companies in general engineering, film processing, clothing, furniture, vehicle assembly, glass, stainless-steel products and pencils; it has investments in firms making aluminium products, hosiery, abrasives, chemicals, stoves and electronics. It has recently invested in joint ventures with local companies (making polypropylene bags for agriculture) and foreign companies (an explosives factory with Swedish interests), and is involved in major new developments which could lead to the production of chemical pulp and paper, caustic soda, plate glass and copper tubes. After the formation of the Zimbabwe Mining Development Corporation (ZMDC), however, it sold its mining interests, including the Kamativi tin mine.
Government has also made direct (usually controlling) investments in CAPS (pharmaceuticals), Zimpapers (publishing), Heinz-Olivine (oils, fats and canned food) and Zimbank, and it has recently acquired a large interest in the largest company of all, Delta Corporation. Most of these expansions of state ownership followed from tactical decisions relating to such factors as the need for cheap drugs for the expanded health service and the reduction of South African influence in key areas such as information and banking; all involved market purchase of shares by the state with no suggestion of expropriation, although government was often able to drive a hard bargain.

State and market
A more formal state role in industrial development was promised in Growth with Equity in 1981 in the proposal for a Zimbabwe Development Corporation, which was not in fact established until 1988. The slow progress in this direction is clearly associated with the more general failure to grapple with the problem of planning for industrial development, itself constrained by the context of continuing private ownership of the majority of industry. Despite the fanfares associated with the Transitional National Development Plan, 1982/83 -1984/85 (TNDP) and the First Five-Year National Development Plan, 1986-1990, no coherent industrialisation plan was included, and no instruments or institutions for such a plan were processed. Many industrial projects were included, but these were largely put in on an ad hoc basis, depending almost entirely on whether they were already being considered by a company or other interest group; there was no prioritisation or attempt to relate them to each other or to other industries needed as part of a planned process of industrialisation.
In fact since independence, state policy towards industry has been very conservative, with the inherited structures of state intervention and protection through control foreign exchange and trade, being preserved with only marginal changes. This may seem paradoxical in that these structures were set up largely as defensive measures against sanctions, and were designed to safeguard the privileges of the white elite. Why, we might ask, should they be preserved by a government inspired by Marxism-Leninism, and seeking greater equality? The answer probably lies in the new state's desire for control over the mainly white or foreign owners of industry, its desire not to risk de-industrialisation by removing protection, and a stated desire to plan the economy. But it is much less clear that instruments to protect a minority regime against sanctions are the best ones to protect and develop an industry for an independent nation. The conservatism has certainly preserved industry and allowed it to grow modestly (the volume of output in 1989 was 31 % higher than in 1980), but there has also, despite the rhetoric, been a conservatism in functional ideology.
I shall try to explain why I say this: we have seen above that before independence industry developed under protection, despite the open-market economy ideology of the colonial and rebel governments. But even after independence, under a nominally Marxist regime, the government -or at any rate the ministry of finance, economic planning and development (MFEPD) -from the start made regular statements that its aim was liberalisation of the foreign exchange and trade regimes, even if in part this may be seen in response to pressure from the World Bank and the IMF. What is clear is that there has been no attempt to develop an alternative to liberalisation through the construction of a coherent plan for deepened industrialisation, although precisely that has been called for by some other sections of government. The controls provided by existing structures have been used primarily to maintain an overall balance of payments, secundarily to sustain and develop export-earning (or importsaving) industries, and only intermittently and on an ad hoc basis have they been used, as they might have been, to pursue longer-term objectives.
It is true that the state role has expanded, both in ownership and control. However, except at the purely rhetorical level, this has not related to any grand socialist strategy or any attempt to introduce central planning: state interventions have always had parallel pragmatic justifications. The confusion between socialist rhetoric and conservative pragmatism may be most clearly seen in the introduction to the TNDP where in one paragraph there is an abrupt transition from a statement of the need to develop socialism and greater equality, to a commitment to promoting the conditions for greater foreign investment, with the clear implication that the authors believe that the latter can promote the former.
Liberalisation also appears to be a theoretical idea for the white dominated business community, with a phased implementation of it gaining verbal support from the Confederation of Zimbabwe Industries (CZI) in 1987. In practice the members of the CZI are loud in their condemnation of particular breaches of protection, and they generally recognise that they benefit from state interventions, not only in providing protection but also (for established firms) reasonably guaranteed foreign-exchange allocations at a favourable exchange rate.
The continuing foreign-exchange shortage and the fear of losing control of the balance of payments therefore prevented any significant moves towards liberalisation on the part of government until this year. Thus we find a continuity between preand post-independence policy applying also at another level: external factors have forced a policy broadly favourable to industrialisation despite the official desire to operate a different policy.
The theme that I shall therefore be developing is that Zimbabwe's policies towards industrialisation have been broadly correct, but inadequately theorised, and pursued with insufficient priority. They have been prey to short-run balance-of-payments constraints which have often resulted in perverse outcomes and serious inefficiencies which could have been avoided. Consequently they have been unable to respond adequately to the continuous sniping from the World Bank from its base of a more coherent (but I believe incorect) orthodox, free-market theory of development.

The World Bank and the Competitiveness of Zimbabwean Industry
The 'Jansen Report'(2) on the international competitiveness of Zimbabwe's industry, was commissioned in 1982 in connection with the first loan that Zimbabwe received from the World Bank; the 'Belli Report'(3) An Industrial Sector Memorandum followed four years later. Now both reports were highly critical of Zimbabwe's past and present industrialisation and foreign-exchange policies, so it is important to appreciate at the outset that even in the reports' own terms the overall results they obtained pronounced a quite favourable verdict on Zimbabwe's achievements: Although Zimbabwe manufacturing may not be clearly efficient in an absolute sense, it is very efficient compared to other manufacturing sectors in Africa and many other developing countries. (4) Whereas the overall DRC(5) was estimated at 1.27, similar studies had given 1.95 for Ghana, 1.83 for Cameroon and 1.34 for Ivory Coast.
As the last-named country is frequently praised by the World Bank for its relatively open policies (6), it should have seemed all the more remarkable that Zimbabwe should show a better result, after nearly two decades with a relatively closed economy, during which it added to its industry a range of more sophisticated processes requiring more protection as infants in any case.
Belli's solution is to express surprise at the success (despite the system), and then^to pass to a catalogue of what it sees as disadvantages of the system (some real, some already solved, some merely disadvantages to the operation of capitalism), without ever mentioning any advantages.(7) All World Bank or IMF reports in the end turn what evidence they have to recommanding the standard free-market package, modified only slightly for a country's specificity.(8) Even if one where to accept that this package was appropriate for a country (like Zambia) in serious trouble after a commodity collapse and failed industrialisation strategy (although this could of course be challenged), it does not necessarily follow that it would be appropriate for a country with a significantly more successful industrial record. One is therefore tempted to ask how successful Zimbabwe's policies would have to be before the World Bank would be obliged to alter its prescription.
My detailed criticisms of 'Jansenism' follow in the next section, and fall under two main headings: misuses of the methodology and neglect of the reservations made by more sophisticated practitioners of the methodology such as Bhagwati. (9) The Jansen Report The Jansen study attempted to survey the ten main branches of manufacturing industry, investigating 122 firms representing 62 % of the output in 1981. (In fact as it restricted attention to only three products per company it covered only a tiny fraction of the 6200 products listed in Products of Zimbabwean Industries in 1980.(10)) The efficiency of these firms relative to international competition was judged by calculating their domestic resource costs. Where these were above unity Zimbabwe would, according to the ideology, gain by closing down such firms and devoting the resources so freed to ones with DRCs below unity.
The methodology of the Jansen Report is thus very simple (or simplistic?); however, it is very difficult to judge the quality of the data inputs.(ll) The report in fact falls short of acceptable standards even in terms of its own methodology. The DRC approach (even ignoring the caveats as to the use to which its results may be put) requires genuine shadow pricing of factor costs, painstaking collection of international prices of tradables, and more reliable calculation of the appropriate exchange rate. If one wished to be cynical one could say that Jansen had considerable difficulty in achieving a DRC above unity for Zimbabwe, and this achievement depended on both overvaluing the social price of domestic factors and undervaluing the social price of value added in the output of Zimbabwean industries.
As countries in early stages of development (Japan and South Korea are probable examples) have often tolerated negative value added in infant industries for moderate periods, with consequent large and even negative DRCs, a DRC value of 4 (as found by Jansen for some metal industries in Ziambabwe) may be seen as quite moderate. And yet this figure is used to imply extreme inefficiency, for it states that $ 8 million of domestic resources are earning or saving only $ 2 million of foreign exchange, whereas if transferred to an activity with DRC of unity they would be earning or saving $ 8 million. Even if we accept the average DRC of 1.27, with most sectors of industry having values between 0.9 and 1.4, a value approaching 4 (with 4.4 for Zisco alone) appears to be damning. This however is an artefact of the way that DRCs are calculated: it can easily be shown(12) that a rise in revenue of only 30 % would suffice to reduce the DRC to unity.
The ideological blinkers of the Jansen Report result in it attributing all variations in DRCs to inefficiencies in the trade regime. Indeed Jansen goes further, at one point(13) even stating that market imperfections are 'considered to be entirely the result of government policy'. Jansen, like the World Bank, advocates a larger role of the market, including the world market; in the context of an 'efficiency analysis', 'market imperfections' becomes a loaded term implying that there is no alternative but a change in government policy in the direction of trade liberalisation.
As the overall DRC for Zimbabwe's manufacturing industry was calculated at only 1.27,(14) it could hardly be claimed that Zimbabwe should not be manufacturing, but should concentrate on mining or agriculture. However, with a spread of 0.66 to 3.62 amongst the branches of manufacturing industry, the possibility of serious inefficiency in some activities is raised. And in some of these branches individual companies scored still higher (up to 5.44) on the DRC scale.
I am not therefore suggesting that such figures (if correctly calculated -see below) should be ignored, but a reasonable approach would seek what lessons were to be drawn in terms of installing more modern plant and equipment, much of which had been run down following 14 years of sanctions. Jansen, however, despite conceding that for the metal products sector 'a major reason for the poor performance of the firms in this activity is the fall in the world prices for steel and metallic minerals' nevertheless concluded that '...the country would save foreign exchange by closing them down instead of running down their fixed capital. The ensuing massive layoffs of workers would be undesirable, but a study of alternative product lines and more efficient use of existing plant and machinery should be undertaken'. When would such a study not be good advice? The important point concerns the long-run prospects for industry rather than short-run savings during a world slump.

The World Bank's Industrial Sector Memorandum: The Belli Report
The Belli Report(15) was the first substantial in-house analysis of Zimbabwean manufacturing industry by the World Bank. As such it is interesting to discover whether it takes account of the almost universal and damning criticism of the Jansen report (privately admitted to have effectively discredited it), which led to the latter's rejection by the Zimbabwe government. Unfortunately once again, although Belli begins with a brief history of Zimbabwe's experience with industrialisation, and discusses the specifics of Zimbabwean practice and institutions, the recommendations which follow are in all respects exactly those which someone with no knowledge of Zimbabwe, but familiarity with the World Bank, would have predicted.
In its consideration of how the present structure of manufacturing industry has arisen, the Belli report acknowledges that the effective start of industrialisation occurred during the Second World War under natural protection and some state initiatives in the steel and textile industries; that the Federation with what are now Zambia and Malawi from 1954-1963 benefitted Zimabwean industry internally through widening its market, and externally through tariff protection; and that the enforced protection during the UDI period resulted in the creation of about half of present industry, protecting some 6000 distinguishable products as compared with some 600 before UDI. What the report plays down(16) is that this was a process involving a high measure of both state protection and state control and initiative (in key respects if not always in detail). It also attempts to play down the importance of the UDI period, when protection and control were most visible, by emphasising that a half of industry was in place beforehand (implying wrongly that this had arisen primarily through market forces)(17), and that structural changes inside the manufacturing sector during UDI were slight. (18) The scientific and objective tone of the Belli report is in fact spurious, for the essence of the scientific method is the rejection or modification of hypotheses in the light of conflicting evidence. However, as in the case of Jansen, the Belli mission seems to have started from a received and inviolable theory, so that the task it set itself became one of fitting facts that could not be ignored into the framework, without modifying it in any way. At its weakest, when it is obliged to report that Zimbabwe has had success with policies of a type that it recommends should be abolished, the report can do no better than state that this success is surprising! Equally, no mention is made of countries which have had failures with World Bank policies, or those which have had success with opposite policies. (19) The propagandiste intent is also evident, despite the sober language, through the fact that only arguments against the present system and its disadvantages are presented. In some cases we may in fact agree with their recommendations, but this would be as a result of a judgement that the disadvantages outweight the advantages; however the latter are never even referred to by Belli, although in my view they are crucial.
As with Jansen the over-riding criterion employed is that of short-run efficiency: present policies are judged inefficient because the market would dictate a different pattern of relative prices and resource allocation. Considerations of equity, of a desire to construct a socialist society, even of the need for structural change for development, are implicitly relegated to a subsidiary category whose main characteristic is that of imposing costs above the level of a market outcome. This implies an underlying assumption that the basic structure of the economy is sound: all that is needed is to allow market forces to favour those actvities for which a comparartive advantage already exists. The report, in other words, denies the need to promote structural change (despite this being an oft-repeated government aim), either for social justice or for the purpose of developing comparative advantage in new areas. The primary aim of the Belli report is thus to accelerate the process of reintegrating the Zimbabwean economy into the world market system in such way as to make any attempt at an independent policy (even of economic nationalism, let alone of socialism) impossible.
As the conceptual and methodological weaknesses of the Jansen report had been severely criticised(20) and as the report was rejected by government, the World Bank understandably wished to distance itself from Jansen; however it clearly wanted to retain the conclusions, and to quote the results of the earlier study where it was unable to repeat the field work.
The Belli Report is therefore by and large only implicitly critical of Jansen (21), despite quoting significantly more favourable DRC values from its own field work; although this changes the specific recommendations, the general ones are untouched. Thus Zisco, which on Jansen's figures should have been written off immediately (22), is now judged to be potentially very efficient with a significant comparative advantage (i.e. DRCs below unity) if appropriate adjustments are made to the crude figures. (23) The report gives a number of reasons why it obtained such a startingly different result from Jansen, including the subsequent devaluation and her failure to calculate proper shadow prices. These confirm some of the earlier criticisms made of Jansen and her uncritical use of the DRC methodology, and in particular the extreme sensitivity of the ratio to changes in one component of the denominator.(24) But if a reinvestigation of Zisco can change a DRC of 4.4 (meaning, let us remember, that for every US$ of value created, US$ 4.4 worth of domestic factors are consumed -i.e. US$ 3.4 are wasted), to a DRC of less than unity, how reliable should we regard her other calculations? And even if they were accurate how useful are they if they are so sensitive to change? And what weight should we attach to her overall DRC figure for manufacturing industry of 1.27? Should it really have been about 0.9? Or 0.5? (Belli's other recalculations relate to only five firms in the textile and fertiliser sub-sectors, and they also are significantly, though not quite so dramatically, below Jansen's). Nevertheless (despite privately expressed embarassment) the Belli report explicitly admits to drawing on the Jansen study, and indeed falls back on it repeatedly.
The Belli Report shares with Jansen an obsession with static comparative advantage (which is what DRCs really measure) to the near exclusion of longer run considerations. It is accepted that it is common for governments to protect industry, but the reasons for this are not discussed. Neither the exception to neoclasical theory of infant-industry protection, nor structuralist, nor socialist theories of industrialisation are mentioned(25). Equally one could not guess that 'model' countries, like Japan and South Korea at early stages of their development (to say nothing of the socialist countries), consciously decided to create new comparative advantages through protection; that is they invested heavily in industries which had high DRCs at the expense of existing ones with lower values. In arguing against the present system, the report implies that it might have been necessary during UDI, but that those conditions no longer obtain. This may well be true, but it misses the point that if the high degree of protection and the discretional (as opposed to market-determined) allocation, were responsible in large measure for the economic success (as we would argue), there is no reason why these elements cannot be used now under more favourable conditions. The boot should really be on the other foot, for it is the report's recommendation that a free-market, export-led startegy could generate the type of success seen in South Korea, that assumes conditions that no longer obtain. Aside from the fact that the internal components of the South Korean strategy were far from free market, few would maintain that the favourable world market conditions enj oyed by that country in the crucial take-off period of the 1960s obtain now, or that (unlike the case with planned allocation mechanisms), developing countries can do anything significant about it.

Jansen and Belli in Perspective
In effect these two reports present the pattern of DRCs amongst branches of manufacturing industry in Zimbabwe as a 'hit list'. The arguments in Jansen amount to the claim that any shift from a high DRC firm to a lower DRC firm (it is irrelevant whether they are both above or below unity, or whatever) will increase efficiency: that is the spread of DRCs is (by definition) too large. Jansen's arguments lead to the conclusion that in the end all resources ougth to be employed in the single activity with lowest DRC. If this ridiculous conclusion (implicit also in Belli) is rejected, where does one stop on the way towards such an outcome? The report offers no guidance; it might be quite soon, or it is even possible that Zimbabwe might have already gone too far towards this extreme. Any position of differential DRCs is open to the arguments they make.
In order to promote international competitiveness the Belli report suggests that either the most capital-intensive new machinery has to be imported, or real wage-rates have to be squeezed down to the levels of the main far Eastern competitors. Both options are inconsistent with Zimbabwe's socialist aims in the short run. Devaluation could increase the competitiveness of the industry, but at the expense of making imports more expensive, reducing living standards, and transferring income from importers and their customers (industrialists and consumers of industrial products) to exporters (largely foreign-owned mining, ranching and plantations companies and settler farmers).
The Jansen and Belli Reports sit so centrally in the mainstream of World Bank and IMF advice that their major procedural failings merge imperceptibly, through the ignoring of Bhagwati's caveats, into the ideologically loaded nature of World Bank methodology in general. The methodology, in subtler hands and with painstaking data collection, can certainly derive reliable knowledge on static comparative advantage (and thereby provide a useful guide to short-run efficiency);(26) there is, however, no methodology that can reliably handle dynamic comparative advantage, dealing as it should with several dimensions of the future: the gaining of economies of scale by infant industries, learning by doing, technological change internally and externally, short and long-term changes in world markets, and so forth.

Zimbabwe's alternative
The Belli Report, recognising that there would be strong political resistance to its proposals, described what it calls 'a second best alternative'(27). This would involve various measures designed to promote exports at roughly the present exchange rate, to reward exporters with forex allocations for domestic-market production, and to make it easier to import essential spare parts and similar commodities. Import, price and investment controls would have to be retained in this case: Belli saw this as a negative factor because of 'the distortions that they introduce' -i.e. relative to the world market prices(28). I would however argue that these controls could be simplified and reformed in several ways, both so as to increase the efficiency of their operation and also so as to allow the introduction of more economic analysis into the decision-making process.
In fact Zimbabwe was already following just such a policy, partly as a result of earlier World Bank advice. Thus it operates an export-revolving fund, in which companies with firm export orders can obtain effectively whatever forex they require in order to purchase essential imported inputs (and in some cases spare parts or even new capital equipment): this was originally funded by the World Bank, but was soon operating from the Reserve Bank of Zimbabwe (RBZ) as part of its normal procedures. There is also an export-incentive scheme, and a domestic-market forex allocation based on the extent to which firms raise export levels. In addition a policy has been followed of managing the exchange rate of the Zimbabwe dollar so that its value as measured by a trade-weighted basket of foreign currencies has been falling slowly in real terms. These measures have raised the volume of exports of manufactured goods every year for the last six years.
Thus Zimbabwe has been following successful policies, unlike most countries in sub-Saharan Africa (including South Africa), as evidenced by an average growth rate in the 1980s of almost 4 %, in which 'non-traditional manufactures' have played a large role. Undoubtedly these policies could still be significantly improved, in terms of raising export earnings and in terms of more rapid growth of industry for the domestic, regional and wider export markets. But I will conclude this section by making two further points: 1. These polices are in essence the same as the policies of the newly industrialised countries (NICs), in particular South Korea, where contrary to orthodox mythology, there was significant government intervention, protection of the domestic market, highly selective quantitative controls on imports, the linking of import licences to export performance, and a battery of other export-promoting measures having little to do with short-run market prices. Zimbabwe has been less succesful, partly because it has had much less consistent or thoroughgoing policies, partly because of a much less favourable geographical situation (it is landlocked, and it has been subject to déstabilisation by South Africa), partly because it received proportionately much less aid and fewer trade concessions, and partly because the world trading environment was much less favourable in the 1980s than in the equivalent period of the 1960s for the NICs.
2. For all Belli's concession that such policies are at least 'second best' (i.e. not totally wrong or counterproductive, merely theoretically sub-optimal according to orthodox ideology), the World Bank has in practice acted to reduce the extent of their success. Thus as the time that the beneficial results of the export-revolving fund were being recorded, Zi mbabwe began negotiati ons with the Bank for a larger fund which would have extended the benefits to the agricultural and mineral exporters also. Such a fund was agreed in principle in negotiations with the Bank's country staff, and was at the point of being signed when it was vetoed by the ideologues at the highest level of the Bank. Four years later it has still not been made available. The price Zimbabwe was being asked (and refused) to pay involved faster devaluation, a commitment to thoroughgoing liberalisation, and, it is believed, an end to socialist rhetoric. As the new fund would have improved an already successful policy, it does not seem too cynical to argue that the Bank refused to support it precisely because it might have succeeded only too well.
Zimbabwe, by then having a good credit rating, was nevertheless able to borrow commercially to set up a similar fund of its own. This of course was on much less favourable terms than the World Bank could have offered, so the outcome is that the latter has in effect deliberately depressed Zimbabwe's exports in pursuit of the imposition of an ideology that argues for the primacy of international trade!

The World Bank victorious?
During 1988 and 1989, with rapidly rising exports and falling debt-service ratio(29), calls for liberalisation lost much of their force. However, it was also clear that the existing system was operating less and less efficiently in that there were increasing delays in making decisions, and increasingly widespread corruption (the evidence from the Sandura Commission(30) was only the tip of a large iceberg) was throwing the rationality of the foreign-exchange allocation system into doubt. In 1988 the widely respected and dynamic permanent secretary in the ministry of trade and commerce was sacked and charged with corruption -on a very minor matter which he claimed was dug up precisely because he had refused to go along with major malpractice by his superiors. The consequence for a time was that almost no decisions were taken on foreign-exchange allocation in the ministry, because more junior staff feared that any discretion could in principle lead to charges of corruption from quarters that felt that they had lost out.
Nevertheless, this was still the system that had delivered positive results, and the most obvious diagnosis should have been for reform. This is what was argued by the ministries of industry and technology, and labour and manpower development. The more orthodox ministry of finance, economic planning and development (MFEPD), however, plainly wished to reach an accomodation with the World Bank: an event seen as highly symbolic by the press -the marriage of the minister's daughter to the son of the World Bank's resident in Zimbabwe -also took place in 1988. The compromise reached seems to be that a phased liberalisation over five years has been agreed on, and may be implemented from July this year. Very careful monitoring will take place, so that firms will not become unviable simply because of earlier inability to retool caused by forex shortage, or because of closure of local sources of supply. Nevertheless the basis does seem to be a switch from quantitative controls to a tariff-based system, which will remove from government a major instrument for planning industrialisation policy.(31) Since, as we have seen, such instruments have not actually been used to a significant extent, it is possible that the net result may yet be beneficial, if control is retained over strategic parameters of policy, including the tariff structure (this will mean resisting World Bank demands for lower and flat-rate tariffs). One pointer to possible disadvantages, however, is seen in the reported 'own goal' by government when it introduced a preliminary liberalisation measure late last year, raising the level above which importers need to seek an import licence from ZW$ 500 to ZW$ 5000. This was intended to ease greatly the obtaining of spare parts by industry. Instead there is evidence of a big increase in the import of luxury consumer goods, especially computers, video recorders and colour TVs, and a halving of the black market value of the Zimbabwe dollar as those with spare local funds try to mobilise the necessary foreign exchange. This may be only a small fortaste of the consequences that opponents of liberalisation have warned of, should the phasing, monitoring, or consequent corrective action prove deficient.

THE FUTURE
Attempts to predict the future almost always fail, even (or especially) when formal mathematical models are available. Intuition (or guesswork) based on detailed knowledge of an economy often provides as good a guide as any, but only if neglected factors do not turn out to be significant. Here I am going to retreat onto the safer ground of scenario construction, based on certain explicit assumptions (for some of which I have no secure grounds). The assumptions are: that there will be no more than one severe drought affecting core agricultural areas every five years (the south and west will probably be affected every other year); that déstabilisation by South Africa will cease to be a significant factor and that most of Zimbabwe's trade will switch to using Mozambican ports.
The two main factors are unpredictable internal and regional policy outcomes, and together they lead to six main scenarios. The first relates to the question of liberalisation of economic policy in Zimbabwe; here it is possible that: 1. government will retreat into perpetuating the present system without substantial reforms; 2. it may reform the system or implement phased partial liberalisation which places control over key economic instruments in the hands of active planners; 3. rapid liberalisation may occur, turning the majority of decisions over to the market.
Regionally, South Africa may: a. reach a settlement in which market policies and an inflow of foreign capital restore it to a major sub-imperialist role; or b. the settlement may be of a nationalistic, even socialist, nature, with South Africa becoming a valued member of the SADCC.
Of the six possible combinations of the above two factors, I will elaborate only 2b and 3a here. 3a, which I shall refer to as the World Bank model, would result in rapid de-industrialisation. 2b, an extension of the existing Zimbabwe model, offers the only significant hope of long-run industrialisation, and therefore of any form of meaningful development.

The World Bank Model
Liberalisation should in theory raise the (static) efficiency of an economy because less efficient firms and activities will be forced to become more efficient or else they will be replaced by more efficient importing or by new firms. Now almost all sub-sectors of manufacturing industry in Zimbabwe contain efficient and inefficient firms, but because of the small size of the economy, many products are produced in a highly monopolistic context. Therefore over-rapid liberalisation would produce rather random results, with lines of production disappearing in favour of imports not because Zimbabwe has no comparative advantage in their production, but rather because they happen to be produced by less efficient firms (or by efficient firms which have not been granted the foreign exchange to invest in modern technology). Initially therefore only some highly efficient sections of industry would certainly survive, but eventually many of these would lose their local linkages, and eventually there would probably be few survivors not tied to mining and agriculture and basic food processing.
With the proximity of an open South African economy of much greater size, rationalisation would reduce the Zimbabwean economy to a periphery of a periphery: relatively few industries, and probably no highly advanced ones, would survive in South Africa itself, and those which did would have a huge competitive advantage over their Zimbabwean equivalents.
Static comparative advantage would be the prime criterion, and the integration of the region into the world economy would force specialisation in primary production: mineral products and a limited range of agricultural crops. Industry which added value to these would survive to some extent, but the most obvious source of comparative advantage -cheap labour -now has no relevance whatsoever to almost all manufacturing industries. It is noteworthy that whereas advocates of liberalisation offer an inflow of foreign investment as a job-creating consequence, very little private investment has flowed of recent years to any part of the Third World, and what has, has been of a remarkably capital-intensive nature. Almost no new investments in Zimbabwe recently (domestic or foreign) have created jobs for less than ZW$ 100,000 each.(32) These are the levels dictated by the need for competitiveness in the world market; a simple calculation shows that even investment at ZW$ 50,000 would only create jobs for a tenth of Zimbabwe"s school-leavers even if the investement ratio rose to over 30 %. The World Bank model is thus a recipe for continuing poverty for a region which has been allocated the role of cheap commodity provision for the developed world, and which would provide remunerative employment for only a fraction of its population.

The Zimbabwe Model
Zimbabwe has industrialised -as almost all countries before it -on the basis of a measure of protection from more advanced economies, whether natural, arising from war or sanctions, or resulting from government policy. Naturally because protection is protection against more efficient firms and economies elsewhere, it is protection to present inefficiency. Preservation of industries because they happen to exist now, or creation of new industries because they are technologically possible (plans are current in Zimbabwe for establishing a machine-tools industry) cannot therefore be justified independent of considerations of future efficiency.
But 'efficency' is a question-begging term. We could easily accept infant-industry protection on a mechanical basis for firms or activities which can demonstrate that they will be mature and internationally competitive in, say, ten years' time; even the World Bank might agree. More important is to look at the efficiency of an economy as a whole: an economy employing barely 10 % of its population (as Zimbabwe is now) and consigning the rest to subsistence and handouts, is neither efficient nor humane. But an economy spreading similar investment over most of its workforce (and thereby raising employment five-fold) would produce almost nothing at internationally competitive rates, and would therefore have to subsidise exports, thereby violating GATT, World Bank and IMF principles. Zimbabwe has not yet been able to follow such a policy very effectively, although since independence it has kept alive the possibility of switching to it, and as we have seen has developed a way of breaking into export markets that could be developed much further (learning from the South Korean experience for example). South Africa will face almost exactly the same dilemma, with the aspirations of its highly politically conscious population certain to be thwarted if World Bank policies are imposed. The sharpness of the contradiction in South Africa could still be resolved in a socialist revolution, but short of this may still allow a compromise or stand-off in which more egalitarian industry-and enployment-generating policies become a possibility. Both South Africa and the SADCC region, including Zimbabwe, could benefit from regional cooperation in this eventuality. There is no space here to spell out the details of corporation and specialisation that would be required (the SADCC provides a model for the very early stages), but I will conclude by making one rash prediction: that the failure of IMF/World Bank structural adjustment policies for other countries will become manifest during the next five years, so that a measure of space may become available for other countries and regions attempting industrialisation and modified structural adjustment on the Zimbabwe model rather than the World Bank one. 14. And as the dollar was probably overvalued by about 30 % in 1981 a more correct average DRC may have been below unity! 15. Belli Report.

There is some evidence of differing views amongst members of the team on a number of issues: in this particular case we find the significance of the pre-UDI period (and its supposed free market character) being argued on pages xv and 71, with this view being explicitly or implicitly contradicted on pages 8, 84 and 96.
18. Belli report, page 12; in fact the Table 1.8 referred to in the text shows that the proportion of total gross output in manufacturing industry accounted for by the metals and metal products branches rose from 14.5 % to 22.1 % between 1967 and 1975, an increase in share of more than 50 %! This is a major change in a key sector for self-sustaining industrialisation, and depended on both state action in underwriting the expansion of the steel industry, and continued protection of a range of metal products industries that few developing countries have yet developed. 25. Protection, however, is seen by at least one member of the team as having played a large role in Zimbabwe's development (pages xiv, 3,4 and 60), but this is contradicted in several other places: see especially page 12.

A more recent World Bank study -'Zimbabwe: private investment and government policy' (1989) -represents an improvement methodologically on Belli, but in the end makes the same recommendations.
27. Belli, page 81.
28. Although the latter do represent short-term opportunity costs for Zimbabwe, they are by no means market equilibrium prices, being heavily influenced by cartellisation, trade agreements and protectionism, and the operation of monopoly power by the TNCs. They will most probably give the wrong signals for a country seeking to change its economic structure and create comparative advantage in new ares.

INDUSTRIALIZATION IN ZIMBABWE: A NON-REPLICABLE MODEL ? INTRODUCTION
Zimbabwe is not just an average African country. In many ways it does not even reflect the conventional picture of a 'Third World Country' (whatever this may be). In many respects the country defies classical development studies labels. Macroeconomic aggregates about its economy often hide more than they reveal. This makes any discussion about Zimbabwe's 10 years of independence a difficult undertaking. I have only worked some five years in Zimbabwe (which is much less than my esteemed colleagues today), and I admit to be intrigued as ever about its economy, politics and policies.
Zimbabwe has by many standards a remarkably diversified economic structure. The country has a sizeable commercial agricultural sector, an important manufacturing sector and a relatively important mining sector (particularly because of its exports)^) In fact this diversity is maintained when looking at a lower level of aggregation. For a tropical country Zimbabwe has a highly diversified agricultural production (major crops are maize, tobacco, cotton, sugarcane, tea and coffee and a vast growing horticulture). In addition, it has a large livestock production both for meat and for dairy products. Last but not least, Zimbabwe is self sufficient in the production of its main staple food, maize.
Its mining sector is not exclusively focussed on one mineral. For sure gold is the dominant mineral but it constitutes less than 50 % of the value of all mineral production. In addition there is coal, asbestos, nickel, copper and chrome ore. Furthermore there is a platinum mine in the pipeline while oil exploration is underway in the Zambesi valley.
Also its manufacturing industry is remarkably diversified. It is not limited to basic consumer goods such as food, drinks and beverages and clothing and footwear but there is also a not insignificant intermediate industrial sector (e.g. chemicals, fertilizers) and small capital goods industry (e.g. metal and metal products and transport assembly).
If we look at the composition of exports, a similar diversity is found: the biggest export earner, tobacco, earns 18 % of total merchandise exports, followed or sometimes surpassed by gold with about the same percentage (18 %). Many developing countries depend often for more than half or even three quarters of their export earnings on one single item, be that coffee, cocoa or copper.
All this diversity with an average per capita gross national income of no more than some ZW$ 480 in 1986.
The other interesting aspect of Zimbabwe's economic history is that a lot of what happened with and to Zimbabwe was not by design of any grand scheme or deliberate economic policy: Cecil Rhodes and his British South African Company actually had a grand scheme about gold but didn't find gold at the scale they hoped for. The attraction of settlers (for small scale mining) and the partitioning out of the occupied lands became the driving force. Industrialization was spurred by another accident, namely the Second World War, although the larger market of the Federation helped it grow. International sanctions gave a second boost to industrialization in the 60s and early 70s and these same sanctions forced a liberal free market UDI government to become very interventionist in assistance to the industry. On the basis of these arguments one might be tempted to conclude that it would be better not to design any grand scheme or policy but leave it to history to be benevolent to Zimbabwe! This, of course, is deliberately overdrawn but it calls for greater modesty in relation to grand schemes and ambitious government policies. To this I will address myself later on.

ZIMBABWE'S FIRST 10 YEARS
The immediate post-independence boom was cut short by a severe drought, poor world economic performance as well as acts of South African destabilization. The average economic growth was no more than 1,6 % per annum over the first five years. The second half of the 1980s gave a more favourable picture. One may achieve an average for the decade as a whole of some 3,4 % of real growth per annum.
Given the high population growth (of nearly 3 %), per capita income declined in the first half, and slighty improved during the second half of the 1980s.
The 'boom-bust' performance of the economy is also reflected in the balance of payments. Its current account deficit worsened in the early and mid 80s only to recover and turn into a surplus in the latter part of the decade. It should be added though that in comparison to the boom-bust fluctuations of other countries those in Zimbabwe are much smaller and influenced by non-economic factors, such as recurrent droughts.
The direction of trade also shifted away from South Africa (the Zimbabwean exports to the RSA were in 1981 22 % of the total exports, in 1987 this had declined to 11 %). Exports moved towards the EEC (from 26 % in 1981 to 37 % in 1987). For imports these changes are less pronounced. The SADCC trade shares have not changed dramatically and have remained well below 10 %.
The foreign debt is very limited. In comparison to other countries Zimbabwe has a low debt service ratio. It reached a peak of 35 % in 1987 but currently it stands at around 20 %.
The most important changes in the economy since Independence have been mentioned already in the previous papers. In agriculture one could see a major increase in the participation of the communal or peasant sector, particularly in commercial crops such as maize and cotton. Up to some 51.000 households were given access to land through resettlement schemes. At the same time, however, the introduction of the agricultural minimum wage together with the earlier drought caused a shake out in commercial farming where employment went down with nearly 60.000 jobs to around 275.000.
In industry there has been even less 'structural change'. The most noteworthy is the relative decline of metal and metal products, and the relative increase in drinks and tobacco and chemical and petroleum products subsectors. On balance the 80s resulted in a 3,6 % increase in real output per annum and a 2,6 % annual increase in employment.
The most important change is found with regard to the services sector: the rapid expansion of education, health and public administration itself. It is also here where the government of Zimbabwe achievements are best known and widely recognized.

Industrial sector
The industrial sector with all its diversity is relatively small. There are less than 1100 units according to the Census of Manufacturing of 1986/7 (was 1340 in 1982). They may produce as mabny as 6000 products but markets are small, and competitors few. The almost completely shielded domestic market generated monopolistic and heavy oligopolistic enterprises which over time diversified into other existing or new product market areas. Many of these economic activities are controlled by a small number of economic conglomerates such as Delta Corporation, the TM group, etc., in which the foreign investment component is still very high. If the commercial farmers form a tight community of 4000 then for sure also the industrialists form an even smaller interrelated community of managers and shareholders.
The most important post-independence feature has been the lack of industrial investment. Gross domestic capital formation (GDCF) was very low, and according to some (e.g. Green & Khadani), there was no net increase in the capital stock. The rate of investment (GFCF/GDP 1980) fell in real terms from 15,3 % in 1980 to an estimated low of 10,7 % by 1989 (in 1975 it was 20 %). It should be noted that the GDCF that did take place in the 80s, was to a large extent undertaken by the public sector, in the form of infrastructural provisions.
Government take-overs via the Industrial Development Corporation have reduced South African shareholder interest, and appears to have been of a defensive nature. Although a small number of new and very large investment projects (e.g. rehabilitaion of ZISCO, the chemical paper and pulp project) are in the pipeline in the form of either joint venture or independent ventures the overall effect so far has been limited.
It is difficult to get a reasonably balanced picture of the major constraints on investment. A recent survey reported the following factors influencing new investment (in order of importance): expected return on investment; forex availibility; insufficient local demand; business confidence (Hawkins et al., 1988).
Whatever the success in Zimbabwe's industrial sector, the fact is that the investment rate has dramatically declined, as I already observed earlier. Formal sector employment has not kept pace with the rapid growth of the urban labour market. If one would make a simple population projection for the 12 main urban areas and if one would extrapolate current formal sector growth taking into account the lower growth rates in the public sector itself (in accordance with existing policies), then the formal sector employment shortfall may be estimated to get to some 620.000 for these 12 urban areas. For Harare alone the shortfall would be in the order of some 310.000 jobs in 1995. This is nearly 80 % of entire formal sector employment in this city (360.000). (2) The unemployment crisis is an urban one and looms heavily on the horizon. The policy dilemmas of Zimbabwe get further complicated. During the recent parliamentary elections ZUM scored its highest success in Harare and it's success is widely regarded as a protest vote. Also one, which I would guess, the government can hardly ignore. The latter means that very expensive investments in urban transport, urban infrastructure and housing will divert the already limited national resources away from the rural areas.
I would submit that the decade of the 80s have represented to some extent a lost opportunity. An opportunity to pursue more vigorously a rural development strategy. The policies in the 1990s are less constrained on the rural side (in the sence that the Lancaster House agreement has expired) but they will have to count more seriously with the more vocal urban problems which also tend to require more expensive solutions.

COMMENTS ON STONEMAN'S PAPER
The industrialization debate as seen by Stoneman is dominated by two major issues: a. the Zimbabwe policy context and b. the World Bank policy advice.

Zimbabwe policy context
With regard to the Zimbabwe policy situation or context, I note some ambivalence in the paper. On the one hand the author asks the question of 'why the Government of Zimbabwe, inspired by marxism-leninism, and seeking greater equality, would want to preserve the inherited structures of state intervention and protection through control of foreign exchange and trade?'. From what I understand of the situation, I think the received view was that these inherited instruments gave the government of Zimbabwe the opportunity to decisively influence the structure of the economy. It gave the government of Zimbabwe awsome powers to control the economy. This was interpreted by some commentators as a very desirable starting point: the government of Zimbabwe had reached the 'commanding heights' of the economy and the socialist experiment could begin.
Later on in this paper the author stresses that the Zimbabwe model has been a correct one, namely, of phased liberalisation with a careful monitoring of developments, to make sure that nobody suffers too much.
I think that the real issue Colin Stoneman raises is not just why preserve such inherited instruments but implicitly he is asking: why were these instruments not really used to effectuate a change away from merely protecting the inherited industrial structure? This leads us to the next key question, namely, doing what instead? What would be a viable long term industrialization strategy? Is there any 'grand scheme' at all?
This same questioning is present in his remarks with regard to the Industrial Development Corporation. The IDC has taken major interests in a number of companies, but for what purpose? What difference does it make?
Foreign exchange allocation has been at the center of attention in industrialization debates. The most surprising aspect of the actual practice of the forex allocation system, which according to almost all observers is the key instrument, is the lack of a clear sectoral prioritization. The major exception to this has been the forex allocation in relation to agriculture and agricultural inputs, where the private sector remained completely involved in the preparation of realistic bids. For all other (non-public sector) sectors the allocations have been based on historic importance and a very general (5-point) product rating as well as general criteria such as 'importance of the rural sector'. (3) This is also evidenced by the fact that cuts in forex allocations, applied in times of extra hardship (e.g. in 1987) were 'across the board' i.e. the same for all applications. The share of 'industrial imports' (i.e. imports by firms except those related to agriculture) fell from 19, 3 % of total forex earnings in 1979 to 8,9 % in 1985. The second development has been the growing importance of barter deals (Eastern European countries and Third World countries) and Commodity Import Programmes (USA -discontinued in 1986 -, Germany, UK, Japan, France, Holland and Scandinavian countries etc.). These so-called ad-hoc allocations constitute between 20-30 % of the total forex allocation to industry. By their very nature CIPs and barter deals cannot be at the core of Zimbabwean economic policy.
In conclusion, it appears to me that even this potentially powerful policy instrument was not applied with any design in mind. It was not given out of hand. On the contrary, access to forex has become more and more an 'art', to put it euphemistically.
It is also remarkable that companies in which the government of Zimbabwe obtained a major or controlling interest were not necessarily better off in terms of access to forex than non-government of Zimbabwe owned companies. At least this is a complaint that was heard in some quarters during interviews in 1987.
The trade liberalization debate, in as far as it is public, suffers from a very similar problem: what sectors will get priority; which sectors will loose out, and what should the tariff structure be like? A phased liberalization looks fine, and indeed better than a big leap in the dark. But...with what priorities?
Perhaps the most important motivation for some form of liberalization may not find its origin in international trade but in the domestic budget deficit. The budget deficit has been growing in the 80s. In 1985 it stood at 10 % of GDP and in 1989 it still remained at 9 % in spite of budget cuts and a healthy economic growth during the last two years. For the current financial year (1989/90) the servicing of debt (interest and loan repayments) stands at more than 1,6 billion dollars which is nearly as much as the combined annual budgets of Ministry of Primary and Secondary Education, Ministry of Agriculture and Ministry of Health together. Further cuts in expenditures are seen as endangering post-independence achievements. The tax base has not increased very much in terms of taxpaying individuals and companies whereas the rates of existing taxes are already considered high. The abolition of quantity restrictions in favour of tariff protection may yield the much needed increase in state revenues. This argument may have relevance but its validity as a motivation for the policy shift towards liberalization cannot be ascertained. It remains sofar a speculation.
Lastly, it is important to make a distinction between 'de-bureaucratization' of forex allocation and 'liberalization' of foreign trade. It would seem that the business community is almost unanimous with regard to 'de-bureaucratization' i.e. simplifying procedures and making them more transparant. The Sandura Commission which exposed corruption among high level civil servants and ministers may also have had an important influence in changing public attitudes towards administrative allocation mechanisms. Liberalizing foreign trade, meaning reducing the protection enjoyed by domestic producers for the domestic market, is quite something else.

World Bank and liberalization
To what extent was liberalization seen as a theoretical ideal for the white dominated business community? I think that here one needs to make a distinction between the first and the second half of the 1980s.
I would submit that neither government of Zimbabwe nor the private sector were in favour of liberalization. Judging from Growth with Equity, Transitional National Development Plan and even the First Five Year National Development Plan, the government of Zimbabwe was not clearly committed to the idea of liberalization. The private sector was during the same period very much on the defensive. The Jansen report was not only controversial in a substantial sense, in particular in as far as the application of the DRC methodology, it was difficult to be accepted not only by the government of Zimbabwe but also by the private sector.
During the 80s the private sector came to realise that in fact the government of Zimbabwe's intentions were indeed 'conservative'. It must have dawned on the private sector that the government of Zimbabwe would not launch a radical strategy but opted for the status quo. However, even under such circumstances the private sector would only be lukewarm about the idea of liberalization. A protected, even if stagnating, market is better than none. A dual strategy in which there is export promotion coupled to forex incentive for the protected domestic market is an attractive option which has now received some cautious support from the private sector. Note that only in 1987 such a global policy gained some verbal support from CZI. Now, in 1990, the government of Zimbabwe is yet to publish details on its liberalization.
It appears indeed correct that, since recently, the government of Zimbabwe is more accomodating with regard to new foreign investment. The creation of the Investment Center and the publication of the 'Promotion of Investment: Policy and Regulations' document in April last year has been heplfull.
Whether the marriage between the daughter of Chidzero and the World Bank Resident Representative's son is symbolic of the accomodation reached between the government of Zimbabwe and the World Bank, I do not know but the fact is that softening of attitudes appears to be mutual. The government of Zimbabwe is more and more underlining the importance of 'an active and viable private sector alongside a strong public sector'(4), the World Bank is currently revising its position and is preparing a new country review.
I would emphasize however that the relations between Zimbabwe and the World Bank when compared with other countries are quite different. Firstly, Zimbabwe has not suffered from the economic collapse which other countries have experienced (e.g. Zambia or Tanzania). Secondly, and for the same reason Zimbabwe has not had a structural adjustment predicament of a scale as Tanzania or Zambia. Thanks to its economic diversity (and not just its conservative policies), Zimbabwe has been in a far better position to whether external shocks.
To what extent is the Zimbabwean model a 'success'? Stoneman does not answer this question clearly. Much would also depend on the criterion used. In the short run, it would appear to be affirmative. Some limited growth has taken place, both in output and even in terms of employment. However, in the long run the declining investment rate is problematic. The outdated capital stock will make new investment demanding more forex. Real wages have been falling below pre-independence levels.
I would add that the Zimbabwean Model has been characterised by indecisiveness between concretizing radical aspirations and pragmatic day-to-day considerations. This indecisiveness may be rooted in many causes, e.g. the Lancaster House constraints, the internal division within the party, the pressure from outside, the conservatism of the bureaucracy, the limited 'technical' planning capacity or implementation bottlenecks etc., but it has had costs in terms of declining investment, a serious unemployment crisis and if judged by the election outcomes also political costs in terms of votes.

Other issues
It appears to me that, relatively speaking, Zimbabwe is faced with important internal policy constraints. These domestic constraints appear to carry greater weight than external policy constraints. In comparison with other less diverse and more dependent countries, Zimbabwe has had more room for developing its own brand of policy. However, internal contradictions and constraints have made it difficult to develop concrete policy alternatives different from the ones currently followed.
A second point that needs to be taken into account concerns the internal dualism between the large peasant sector and the white dominated agro and industrial complex. A dualism that reveals itself on the demand and on the supply side.
The demand side issue reminds me of the Latin American industrialization debate. There, import substituting industrialization had failed to develop the second and third stage of the import substitution strategy. In stead of deepening into more intermediate and capital goods production, industries moved increasingly to smaller but growing and profitable markets for high income products. The highly skewed and worsening distribution of income favoured this trend. Over time industries became less low-income demand oriented. The more this happened, the more difficult it would be to generate an effective policy shift in favour of low income demand (e.g. Leff, 1982)(5).
Zimbabwe has a very similar problem: the per capita expenditure of households of main socio-economic categories is revealing in this respect: The actual expenditures in the last category are nearly 19 times the average in communal areas. There is no evidence to suggest that since 1985 these gaps have narrowed, in stead there are indications that they may have widened.
The Latin American response has been a sidestep: namely export promotion. Can Zimbabwe follow this and is this desirable? There is one important handicap: Zimbabwe is a landlocked country with a small domestic market and hence less capable of developing the Latin American answer of export diversification. Would it not be desirable to expand the domestic market via a redistribution of (underutilised) assets such as land?
The dualism on the supply side is equally serious. There is a big gap between the white dominated industrial sector and the black dominated rural industries. The promotion of small scale industrialization and advancement of black entrepreneurship (to be distinguished from 'emergent briefcase' businessmen!) have met with very lukewarm government of Zimbabwe support to put it very positively. Small scale and new enterprises have suffered much more than large and established enterprises from access to forex and problems with the administrative forex allocation system. (6) There is a widespread agreement now that the expansion of the formal sector, in particular the manufacturing sector, cannot make more than a dent in the unemploy-ment problem. In 1988 the investment required for a single manufacturing job was estimated to be ZW$ 49.000 in 1985 prices (Hawkins et al.). In his paper Stoneman states that the investment requirements have now risen to at least ZW$ 100.000 per job. Staggering investment sums would be required to provide full formal sector employment. Let us be reminded in this context that the per capita annual consumption expenditures in the communal areas were ZW$ 236 (in 1985).
A very practical example could be to respond to the rapid urbanization and consequent need for housing in all market segments with a policy of promoting the construction and construction materials industry, both formal and informal, which generally has a low import content, and is labour intensive. In fact the construction industry itself is currently the sector with the highest employment growth, but suffers from shortages (surprisingly, cement is being exported to South Africa).
The same applies to the horticulture and agro-processing sector. The Agricultural Development Authority is beginning to develop smallholder schemes in fruits and vegetables, with the financial assistance of the EEC.
Perhaps Zimbabwe would not be better off with a grand scheme. I would agree with Colin Stoneman that in Zimbabwe there were grand and radical aspirations and, at the same time, a lot of conservative day-to-day management of the economy. However, this may also be indicative of the difficulty to translate grand schemes into practical and concrete (and well founded) policies. I was somewhat surprised by the rather strong dismissal by Stoneman of the TNDP and the lack of references to the other major 'Plan Documents'. The point is that without these there are very few policy documents left.

Colin STONEMAN
I may have given a false impression when talking about export promotion. By this I was not suggesting the type of export oriented industrialisation policy, whereby the export market is the motor of industrialisation and development. I am not in favour of that model, that has been chosen by certain Latin American countries. I was talking about promotion of export in a purely instrumental way: there is a critical shortage of foreign exchange for the development of the internal market, for the expansion of industrialisation, so as to bring in the rest of the population that is outside the industrial enclave for the moment. It is, like it or not, to generate the foreign exchange so as to back up this industrialisation process. Foreign exchange is easily identifiable as the key constraint on further investment and therefore export has to be promoted. That is a different thing from advocating export-led growth.

Doeke FABER
You didn't go into the possibilities that in the future the SADCC might well provide the possibility for expanded industrialisation: have you thought on that ?
If you know that an additional workplace would cost ZW $ 100.000, have you not thought about a different alternative that would be able to absorb these people that are added every year to the labour force. It is nice to look at what has happened, but if you look at the ways Zimbabwe has to follow, you have not really come up with any realistic avenues along which Zimbabwe could develop.

Colin STONEMAN
Regarding the SADCC, it is fairly easy to say what should happen in principle: there are publications by the SADCC about cooperation between the countries so that they should specialise in different activities, including inside the industry.
Just one example for the textile industry. All textile would be processed on the basis of Angolan oil, which would be polymerised in a plant in Zambia and spun in other plants in Tanzania, Zambia and Zimbabwe and finally made into textiles and cloths in other factories in all countries of the region. That looked a coherent plan in terms of the availibility of resources and the relative need for the number of plants: you need one polymerisation plant, and there were reasons for siting it in Zambia.
That's the sort of logical answer that the SADCC hoped it would be able to give. But there are too many problems for making that sort of regional planning realistic, associated with nationalism. Zimbabwe really doesn't see why Zambia should have these plants when it thinks it can build and operate them better. Even if the plants are built and if the size of the SADCC market says there is only room for one plant in the region, what happens? The other countries that don't have the plants need foreign exchange to buy the output with and so you get into the question to ensure that they do have the foreign exchange. In practice, the plant gets built with foreign aid and then only produces for the country that it is built in, at 20 % output, very inefficiently.
Again, I am listing the problems rather than giving the answers: the answer is easy on paper, but very difficult to achieve in practice. One of the problems is that a lot of things need to happen at once: you need to solve the problem of how to trade by an increased range of export revolving funds and while at the same time making the first moves towards a regional currency. But all these things depend on each other. In principle it might be achievable and with a liberated South Africa it might be easier. We are talking about South Africa as a dynamic member of a coordinating body like the SADCC, not as a member of a free trade area, which might have desastrous consequences for some of the other countries.
The short answer to your other question about what to do to raise employment levels. Basically what I did point to, although too briefly in my paper, was the impossibility of investing enough to create the number of jobs needed. Even if this investment ratio was raised to 20 % or 30 % or even to 40 %, it would be very easy to show that -at a cost of ZW $ 100.000 per job -you would only be creating jobs for a quarter of the school leavers. This means that you would have to get into rural industrialisation, the type of solutions that Dr. Helmsing is talking about. You have to invest in labour intensive jobs and I don't believe that this can happen on a market basis. This must happen with state suppport -e.g. through inputs provision like providing standard metal shapes fore village blacksmiths -and than the output for a long period is not going to be internationally competitive. This is why I say a measure of protection does have to be maintained and probably is best maintained by controls rather than by tariffs. If you can get the cost per job down to ZW $ 10.000, which is still a lot of money, this is going to bring an amazing increase in productivity but it is at the cost of the protection of the economy as a whole. Alongside that, there has to be an internationally competitive second leg, as I was indicating earlier, to export so as to provide the necessary foreign exchange for essential inputs.

François WATERKEYN
You say that trade with the PTA is very marginal compared with the trade with the other countries, especially with the North, but if you look at a field like transport, where I am in, you will see that, when Zimbabwe can compete freely in international bids, e.g. in the field of wagons, in Uganda, Tanzania, it gets all the orders. It means Zimbabwean companies are highly competitive for bids where there is financing, e.g. from the EC, the World Bank or another international organisation. The PTA trade could be improved also in other areas if the financial flows could flow freely within those countries.

Colin STONEMAN
In practice, the PTA as an organisation has not been very succesfull. Those major wagon orders were highly exceptional and depending on aid funding for their financing. I don't think that it depended on the importing countries being members of the PTA, it was a straight commercial deal.

François WATERKEYN
They were competing with many Western European companies: they were more competitive because of the transport cost and because of other factors, but they were competitive. This means that an export oriented growth can be sustained even if they have to compete with European companies -maybe not with South African companies, because these would also have the advantage of transport -, provided the financial flows can be channelled through different countries in Africa and the PTA region.