DEPARTMENT OF AGRICULTURAL ECONOMICS AND AGRIBUSINESS UNIVERSITY OF GHANA AND CENTER FOR DEVELOPMENT RESEARCH (Bonn, Germany) Performance and Problems of Enterprises in Ghana Survey of Formal Private Sector Enterprises in the Agricultural and Manufacturing Sectors January 2004 Dr. Susanna Wolf University of Ghana Department of Agricultural Economics and Agribusiness P.O. Box LG 25, Legon, Ghana Phone: 024-51 48 13 Email: suswolf@web.de We wish to acknowledge financial support by the Alexander von Humboldt-Foundation (Germany) for funding the stay of Dr. Susanna Wolf at the University of Ghana. We also want to thank the survey team members Dr. Daniel Sarpong, Nicholas Nikoi Abossey, Theodora Tettey, and Richard Tweneboah-Kodua for their commitment and especially the participating firms, their managers and employees for their willingness to support this survey.
Table of Contents Summary of findings 1. Introduction 2. Macroeconomic background 3. The enterprise sample Sample structure and representativeness Selected firm characteristics Import Dependency Capacity utilisation 4. Characteristics of successful exporters 5. Investment behaviour 6. Performance indicators for private enterprises Employment growth Development of output Financial performance indicators Firm Productivity 7. Main business obstacles General business obstacles Delays of reforms in the energy and water sector Lack of external support 8. Conclusions and Policy Recommendations List of Abbreviations Literature List of Tables Table 1: Enterprises by size and sector (2000), number of enterprises per category Table 2: Selected firm characteristics by firm size (in %), 2002 Table 3: Selected Firm Characteristics by Sector (in %), 2002 Table 4: Imported Inputs as percentage of total inputs, average 2000-2002 Table 5: Capacity Utilisation by sector and size (in %), average 2000-2002 Table 6: Propensity to Export and Export Intensity(%) by Size and Sector, average 2000-2002 Table 7: Propensity to Invest and Investment Rates (%) by Size and Sector, av. 2000-2002 Table 8: Export and investment categories, average 2000-2002 Table 9: Change in employment by size and sector between 2000 and 2002 Table 10: Real output growth by size and sector between 200 and 2002 Table 11: Rate of return by sector, 2000 to 2002 Table 12: Profit Margin by sector, 2000 to 2002 Table 13: Estimation of Production Functions Table 14: Share of costs for water and energy by sector (%), 2000-2002 List of Figures Figure 1: Macroeconomic indicators, 1990-2002 Figure 2: Non-traditional exports by sector, 1986-2002 Figure 3: Main business obstacles, 2003 Figure 4: Characteristics of enterprises that are using external support, averages 2000-2002 1
Summary Although in general the business environment has improved over the past decade and especially since 1983, the period 2000 to 2002/2003 was not an easy time for manufacturing and agricultural firms in Ghana. On average a decline in the number of employees can be observed. Between 2000 and 2002 the total number of employees in the sample firms declined by 4.3%. In 2003 the number started to increase again but average employment is still lower than in 2000. In line with this more than 50% of the firms were expecting their labour force to increase after 2003, whereas only 12% expected a decrease. The percentage of small and medium firms that expected an increase in their workforce was higher than average and more large enterprises than on average expected a decrease in employment. Wages differ largely by size and sector. On average wages are three times higher in large enterprises than in small or medium enterprises. Contrary to the developments in the number of workers, real output of the sample enterprises increased considerably from 2000 to 2002, indicating an improvement in labour productivity. Especially sectors that face relatively little external competition and where demand increased due to an increase in income were able to increase their output considerably. But for other sectors, such as canned tuna, competition in export markets has increased significantly so Ghanaian enterprises were outcompeted. Of specific importance for the development of the private sector is the development of exports that reflects competitiveness and the investment behavior that lays the foundation for further growth. The export orientation of a firm is higher for bigger firms, younger firms and firms with less foreign participation. However, if small firms export they tend to export a higher fraction of their production than larger firms. The share of raw materials used for production that are imported is relatively high with 50%, which is caused by a lack of adequate intermediary inputs in Ghana. This share is much higher for larger enterprises. Likewise enterprises with foreign participation tend to import more. Less than half of the enterprises invest in a given year. In general the bigger enterprises and enterprises with foreign ownership invest more regularly. The average amount of new investment (in real terms) declined between 2000 and 2002. However, in 2003 it increased again considerably, above the 2000 level. The differences in investment behaviour between exporting and non-exporting enterprises in the sample are relatively small. However, more of the non-exporting enterprises also don t invest and more of the non-investing enterprises also don t export. Exporting enterprises are also more likely to invest. Besides some problems with finding skilled labour and the attitute of the workforce, high interest rates and difficult access to credit are still the main business obstacles, followed by high depreciation and inflation rates and tax regulations. The low quality of water and electricity supply also causes severe problems for the manufacturing sector, as they cause interruption of the production and bind capital for generators and destroyed machinery. Although the government has proclaimed a Golden Age of Business and created a Ministry for Private Sector Development, many enterprises do not regard government policies as helpful and state that the business environment has not improved. Despite the problems and obstacles the prospects for the private sector in Ghana are fairly good. Especially in the horticulture and agricultural processing sectors further growth of exports can be expected. For the plastic and aluminum manufactures who serve mainly the domestic and regional markets competitiveness is also relatively high. 2
1. Introduction Ghana has a relatively good international reputation with respect to political stability and macroeconomic reforms. However, its success with developing the private sector and attracting investment has at best been mixed. For more than a decade private sector development has been high on the agenda of policy makers in Ghana. The aim is not only to foster growth but also to reduce poverty especially through the creation of employment. After initial successes of the structural reform programs between 1985 and 1995, however, especially in the second half of the 1990s, performance of the manufacturing sector declined. A new survey on the performance and problems of private enterprises in the formal commercial agricultural and manufacturing sector in Ghana - which covered the period from 2000 up to 2002/2003 - was used to analyse recent developments in enterprise performance and to find out what measures are required to support them. In the following analysis mainly descriptive statistics are used to link the performance of enterprises in terms of exporting, investment, employment and output growth as well as financial performance to various firm characteristics. To quantify these relationships correlation coefficients are calculated and some simple regression analysis is carried out. From these observations recommendations on how to improve the business environment are made. These recommendations are addressed towards policy makers, donor agencies and the business community itself. 2. Macroeconomic background The formal private sector in Ghana is not yet much developed. Only 7.5% of the total labour force were employed in formal private enterprises in 1998/99 and this percentage has been relatively stable over the past decade (Ghana Statistical Service, 2000). A similar percentage is employed by the government including state owned enterprises (SOE). In contrast 27% of the labour force are self-employed in non-agricultural activities and 56% are in farming which is mainly small scale and often informal. This phenomenon is generally described as the missing middle with a large number of micro-enterprises, some large enterprises but relatively few small and medium size enterprises. Excessive regulation, distorted incentives and state domination of industry are among the reasons for this development. However, the adjustment policies, that were started in 1984, shifted the advantage away from larger firms towards small and medium sized firms that are more flexible to react to changes and entered profitable niche markets (Steel and Webster, 1992). In terms of its contribution to GDP the formal private sector is somewhat more important than with respect to employment. In 2001 industry contributed 25% to GDP. Manufacturing had the biggest share with 9% followed by construction. However, the industrial sector includes water and energy with 2.7% of GDP as well as other SOEs. Growth of the manufacturing sector has been in line with total GDP growth. This growth was partly driven by an expansion of non-traditional exports that picked up again in 2001 and 2002 after a stagnation from 1998 to 2000. Historically agriculture has had the largest share in total GDP with cocoa being the major export crop of the country for the last century. Even compared to other Sub-Saharan African countries the share of agricultural value added in total GDP is high (with 36%) despite its low productivity. Meanwhile the service sector has overtaken agriculture and contributes 39% to GDP. Industry, including manufacturing, with 25% of GDP is still the smallest sector (see World Bank, 2002). The decline in Ghana s terms of trade at the end of the 1990s further deteriorated the fragile macro-economic situation. The trade deficit peaked in 1999 but was reduced thereafter by a 3
sharp depreciation. Ghana s trade regime has been liberalised considerably over the past decades. In general a high percentage of manufactured consumer goods as well as industrial raw materials in Ghana are imported. Although the reduction of inflation is on the agenda for quite some time inflation has been fluctuating significantly up to now (see Figure 1). In early 2003 it increased again due to the increase in petroleum and utility prices. Another issue that limits private sector performance is the low quality of infrastructure. Figure 1: Macroeconomic indicators, 1990-2002 160,00 140,00 120,00 100,00 80,00 60,00 40,00 import taxes (% of total revenue and grants) real effective exchange rate inflation Exports (% of GDP) GDP growth % 20,00 0,00 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Source: World Bank, 2002. Overall the investment climate has been constantly improving over the past decade. Literacy and access to telecommunication are constantly increasing thereby creating the necessary preconditions for investment. In addition the share of the population living below the poverty line has been reduced by 10 % in the past decade. Figure 2: Non-traditional exports by sector, value in USD 1986-2002 600 500 400 300 200 100 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Total agricultural products fish and seafood processed and semi-processed wood products prepared foods/beverages other processed and semi-processed products handicrafts Source: GEPC, 2003 After difficult times for business in 1999 and 2000 because of macroeconomic instability (especially high inflation and massive depreciation) the situation improved in 2001. The Cedi stabilised but the huge depreciation in 2000 lead to a depreciation of the real exchange rate (see Figure 1), which means that exports became more competitive, whereas imports became relatively more expensive than domestically produced goods (Wolf, 2003). However, in 2003 4
the inflation was higher than the depreciation of the Cedi which might lead to some problems for exporters again. However, bank lending to the private sector hasn t improved much in real terms (that is when inflation is deducted) over the past three years. In 2002 almost 20% of bank credit to the private sector went into manufacturing and only 11% to agriculture, forestry and fishing. Most of it went into services (see in IMF, 2003, Tables II.7 and II.8). The main sub-sectors in manufacturing in Ghana are food processing and beverages, textiles and garments, wood products, paper products and printing, aluminum products, plastic products and handicrafts. Most of these sectors also contribute significantly to non-traditional exports (see Figure 2). 3. The enterprise sample Sample structure and representativeness As the focus of the study are investment decisions (foreign vs. domestic) and export performance, a stratified sampling approach was used. 1 The main lists from which enterprises were drawn are the list of non-traditional exporters in 2002 from GEPC and the list of agricultural and manufacturing companies with foreign investment between 1994 to 2000 from GIPC. In addition all enterprises that belong to the Ghana Club 100 (the 100 biggest enterprises in Ghana in 1999 measured by total turnover) and that are in the manufacturing sector were included. 2 The reliance on these lists means that only formally registered enterprises are included in the sample. Further criteria for the inclusion in the sample were: - minimum of 5 employees to exclude self-employed. This restricts the number of agricultural enterprises considerably as most agricultural production in Ghana is done on family farms and in an informal way. - location in the Greater Accra region (where 80% of manufacturing enterprises are located) or Tamale (to capture an area in the Northern part of Ghana with no access to sea-ports). A few enterprises have their headquarters in Accra but their main production site is in other regions so some of the geographical diversity of Ghana is captured. - company exists at least since 2000 to ensure that meaningful data for output exist for at least 2 years. - the company should not mainly be owned by government as the focus of the study is on private sector development. From the combined list of approximately 1000 enterprises a random sample of 300 enterprises was drawn. However, due to time restrictions, errors in company addresses and unwillingness to participate in the survey only 100 questionnaires could be included in the data-set. To capture dynamic aspects of firm performance most questions were asked for the years 2000, 2001 and 2002. However, for the year 2000 the response rate was significantly lower than for the other years, mainly due to a lack of records. The firms in the sample cover most sectors that are relevant in Ghanaian manufacturing and exporting non-traditional agricultural products. They also cover a broad range of firm sizes. (see Table 1). 1 2 A stratified sample is appropriate if firms within groups are relatively similar but there are significant differences between groups, especially size categories. In mid 2003 no current list of all agricultural and manufacturing enterprises in Ghana existed. 5
Table 1: Enterprises by size and sector (2000), number of enterprises per category size category (number of employees) sector less than 20 20 to 75 more than 75 all horticultural 3 3 5 11 fish 0 1 4 5 other agricultural 3 2 1 6 agric. processing 4 7 12 23 textiles/garments 2 3 3 8 wood and furniture 6 1 3 10 aluminium 4 1 6 11 plastic and rubber 1 2 8 11 other manufacturing 2 2 2 6 handicrafts 5 1 0 6 all 30 23 44 97 A small firm is one with less than 20 employees, a medium has from 20 to 75 employees and a large firm has more than 75 employees. 3 Of the 97 firms 65 were located in Accra (all sectors except fishing), 13 in Tema (all fishing enterprises and some agricultural processing, aluminium and textiles), 11 in the Northern Region (mainly agricultural, agricultural processing and textiles) and 8 in other regions. Selected firm characteristics Firm characteristics of ownership and age (number of years the firm exists) vary considerably across sectors and firm size. Most of the firms are limited liability enterprises although smaller firms are also often sole proprietorships. State ownership occurs at all firm sizes but is much more prevalent for bigger firms. These are often former state-owned enterprises that have been partly privatised, often with participation by foreign private investors. Foreign ownership also increases with size and the majority of bigger enterprises has some form of foreign ownership. However, if a firm has some foreign ownership its share in total capital declines with size (see Table 2). Table 2: Selected firm characteristics by firm size (in % of all firms), 2002 size category (number of employees) less than 20 20 to 75 more than 75 All Number of firms 25 28 42 95 Firm Ownership a fully Ghanaian private 74 66 31 53 any state ownership 2 6 20 11 any foreign ownership 26 30 65 44 % of foreign ownership, if any 82 74 68 71 Firm Age in years less than 6 years 37 36 7 24 6 to 10 years 40 27 20 28 11 to 15 years 10 9 17 13 16 to 25 years 6 19 15 12 more than 25 years 7 9 41 23 Notes: a - the ownership categories don t add up to 100 as some firms have foreign owners and are partly state owned. 3 This classification is in line with Rankin et. al. 2002. The number of employees refers to the year 2000. 6
In several sectors such as other agricultural production, aluminium and handicrafts there is no state ownership at all in the current sample. In fishing and agricultural processing around 20% of the firms are partly owned by the state. In all sectors except handicrafts there are firms with foreign investors. In fishing, plastic and rubber production and agricultural processing more than 50% of all firms have some form of foreign ownership. In the wood and furniture sector the share of firms with foreign participation is lowest, but if there is foreign participation it reaches almost 100%. In general foreign investment contributes more than 50% of investment once a foreign owner is involved (see Table 3). That means if foreign investors get involved they want to control operations to a large extent. More than half of the firms were existing for not more than 10 years in 2000. Larger firms tend to be older. 41% of large firms are older than 25 years, compared to only 7% of small firms. In contrast almost 80% of small firms do exist for not more than 10 years (see Table 2). This might be the result of small firms either going out of business or growing into the larger size categories with time. Table 3: Selected Firm Characteristics by Sector (in %), 2002 horticultural other agricultural agric. processing wood/ furniture plastic and rubber other manu facturing textiles/ garments aluminum handicrafts fish Number of firms 11 5 6 23 8 10 11 11 6 6 Firm Ownership a fully Ghanaian private 55 20 67 39 75 70 64 27 33 100 any state ownership 9 20 0 22 13 10 0 9 17 0 any foreign ownership 45 80 33 57 25 20 36 73 50 0 % of foreign ownership, if any 54 68 58 64 60 98 70 91 82 0 Firm Age in years less than 6 years 27 20 17 23 29 10 22 36 33 17 6 to 10 years 55 40 17 32 14 30 22 9 17 33 11 to 15 years 9 20 33 9 0 10 22 9 17 17 16 to 25 years 9 20 17 5 29 30 11 0 0 33 more than 25 years 0 0 17 32 29 20 22 45 33 0 average age 6 8 12 17 16 15 14 15 15 16 Notes: a - the ownership categories don t add up to 100 as some firms have foreign owners and are partly state owned. Firms in the plastic and rubber sector tend to be either older than 25 years (45%) or very young (36%). In other manufacturing and agricultural processing there is also a relatively high proportion of old firms and the latter are oldest on average. In contrast most of the agricultural and fishing enterprises as well as handicraft producers are not older than 15 years. In the wood and furniture and aluminium sectors firms of all ages are represented (see Table 3). Import dependency On average, 42.5% of all raw materials used for production in Ghana are imported (see Table 4). This share is much higher for larger enterprises. Likewise enterprises with foreign participation tend to import more. In the plastic and rubber sector where petroleum based products are the main input almost all raw materials are imported, which makes the sector particularly vulnerable to exchange rate depreciation. On the other end of the spectrum is the wood and furniture sector where less than 10% of all raw materials are imported. In the handicraft sector imports are also low and they are completely sourced in neighbouring countries. In the agricultural sectors the percentage of imported inputs is relatively high as not only fertilisers and other chemicals but also packaging material is often imported by export 7
oriented producers. The import shares remained relatively constant over years 2000 to 2002. However, for aluminium producers it started to increase in 2003 due to the close down of VALCO. Table 4: Imported Inputs as percentage of total inputs, average 2000-2002 size category (number of employees) Sector less than 20 20 to 75 more than 75 All horticultural 11.1 40.5 62.0 38.8 fish - 41.7 44.3 43.8 other agricultural 40.0 52.9 85.0 52.5 agric. processing 34.0 9.0 54.8 38.5 textiles/garments 24.0 31.5 78.0 47.4 wood and furniture 0.0 0.0 40.7 9.5 aluminium 21.0 28.6 30.7 27.3 plastic and rubber 100.0 94.3 99.4 98.3 other manufacturing 83.3 58.0 15.0 51.8 handicrafts 22.3 6.0-17.8 All 26.3 31.8 59.5 42.5 Capacity Utilisation Since the beginning of the 1980s when capacity utilisation in industry was around 20%, the situation has improved considerably. On average the capacity utilisation in the sample enterprises was 63% (see Table 5). From 2000 to 2002 it increased from 60% to 65%. In general capacity utilisation was higher in the larger companies but in the plastic and rubber sector capacity utilisation was highest in the small and medium enterprises. It is also the sector with the highest average rate of capacity utilisation of 80%. In the fishing, other agricultural production, agricultural processing and textiles and garments sectors capacity utilisation was between 50 and 60%. However in all of these sectors capacity utilisation increased from 2000 to 2002 to more than 60%. Table 5: Capacity Utilisation by sector and size (in %), average 2000-2002 size category (number of employees) Sector less than 20 20 to 75 more than 75 All horticultural 56.1 70.0 62.3 62.8 fish - 53.3 50.0 52.0 other agricultural 75.0 50.8 18.3 52.5 agric. processing 34.5 48.3 71.1 57.9 textiles/garments 56.0 66.0 54.9 58.5 wood and furniture 64.6 65.7 85.7 70.2 aluminium 57.3 46.4 68.5 60.6 plastic and rubber 100.0 100.0 72.7 80.5 other manufacturing 57.5 72.5 61.2 62.6 handicrafts 50.0 80.0-65.0 All 57.5 61.8 67.1 62.8 4. Characteristics of successful exporters With respect to exporting, two different trends are observable. Firms that previously only served the local market start to penetrate the ECOWAS market. As Ghana is surrounded by countries that belong to the CFA zone they benefited from the stability of the CFA (which has a fixed exchange rate against the Euro) especially in times of high Cedi depreciation as in 8
2000. On the other hand companies that exported almost all of their production also started to service the local market, partly due to increasing competition in export markets but also to diversify their markets. In total 63% of the firms in the sample are engaged in exporting. This clearly is a higher percentage than the average for all firms in Ghana due to our sampling procedure. In total GEPC lists only 215 manufacturing enterprises, 27 handicraft producers and 178 agricultural producers that were actively exporting in 2002 (GEPC, 2003). On average 36% of all exports are targeted towards other ECOWAS countries. However, it is likely that a significant proportion of exports towards other countries in the region is done by informal traders and therefore not reported. Table 6: Propensity to Export and Export Intensity (%) by Size and Sector, average 2000-2002 size category (number of employees) Propensity to export a less than 20 20 to 75 more than 75 All horticultural 1.00 1.00 1.00 1.00 fish NA 1.00 0.75 0.80 other agricultural 1.00 0.57 0.33 0.72 agricultural processing 0.20 0.55 0.64 0.55 textiles/garments 0.20 1.00 1.00 0.83 wood and furniture 0.00 0.14 0.71 0.20 aluminium 0.40 0.57 0.56 0.52 plastic and rubber 0.50 0.57 0.92 0.82 other manufacturing 0.00 0.20 0.50 0.24 handicrafts 0.77 1.00 NA 0.83 All 0.44 0.65 0.74 0.63 Export intensity if a firm exports b horticultural 83.3 56.2 86.6 75.1 fish NA 93.3 80.0 83.3 other agricultural 100.0 100.0 50.0 96.2 agric. processing 60.0 67.8 50.8 56.2 textiles/garments 20.0 90.0 33.0 60.9 wood and furniture NA 10.0 62.6 53.8 aluminium 60.5 48.3 18.8 35.5 plastic and rubber 5.0 5.0 33.2 28.0 other manufacturing NA 2.0 73.7 55.8 handicrafts 60.0 84.0 NA 68.0 All 72.5 66.7 49.8 59.1 Notes: a - export propensity means the likelihood that a firm in a specific category exports, if it is 1.00 all firms in this category export, if it is 0.50 half of all firms export and if it is 0.00 no firm exports. b - percent of output that is exported if a firm exports. NA means there are no firms in this category that have answered the respective question, e.g. in the fishing industry no small firms were included in the sample. A clearly positive relation exists between size and the likelihood to export as can be seen from Table 6. Less than half of the small firms are engaged in exporting but almost 75% of the large firms export. However, if small firms export they tend to export a higher fraction of their production than larger firms. On average exporting firms export around half of their output, which is comparable to other studies (Rankin et al. 2002). The positive relationship between size and exporting could be explained by different factors. Small exporters have a relatively high share of exports in total output. This observation is compatible with the high fixed costs of exporting, e.g. establishing contacts, acquiring market information or specific product design for overseas markets. Therefore for smaller firms exports are only profitable if 9
a large proportion of output is exported. On the other hand larger firms may be better able to finance these costs and are therefore more likely to export. The huge differences in export shares of the different sectors are due to various factors. For some sectors, transport costs are prohibitively high so they only can export to neighbouring countries (e.g. mattresses and beverages). This explains the relatively small percentage of exports in the plastic and rubber sector. In contrast all horticultural enterprises in the sample produce to some extent for the export market and almost all in the fishing industry, textiles and garments and handicrafts production. In agricultural processing, aluminium production and other manufacturing only around half of the firms or less export. Aluminium products that consist mainly of construction materials (roofing sheets, louvre blades and cooking utensils) are mainly exported to the regional market and therefore the percentage of exports in output is also relatively low. In all the agricultural sectors, the exporting firms export a high percentage of their output towards Europe which indicates a high degree of specialisation among the commercial agricultural enterprises. For example, pineapple growers mainly target the export market and only sell products that don t meet the quality standards on the domestic market. On the average older enterprises are less likely to export. One reason for that is that these enterprises were set up with the aim of import substitution and therefore have to undergo major re-orientation before they are able to export. Furthermore many of the oldest enterprises are subsidiaries of multinationals and often don t have an interest to export as other markets are served by other subsidiaries. On the contrary one cosmetics company stopped production in Ghana completely after it was taken over by a multinational company and only distributes imported products now. In general firms with some foreign participation tend to export less, which means that their main reason for producing in Ghana is not so much exploiting the Ghanaian comparative advantages such as cheap labour but rather to penetrate the Ghanaian market. This corresponds with the relatively high percentage of foreign investors in the agricultural processing and plastics sector which both serve mainly the domestic and regional market. 5. Investment behaviour Most of the 20 biggest enterprises in Ghana s manufacturing sector are foreign owned either by multinational companies or by individuals. In general the bigger enterprises invest more regularly and as they are foreign owned, access to capital is less problematic for them compared to domestic companies. However, due to loans denominated in foreign currencies, some firms experienced high exchange rate losses especially in 2000 when the Cedi depreciated heavily. Overall only 41% of the enterprises did any new investment in a given year (see Table 7), an observation that is in line with previous findings of private firms in Ghana (Rankin et al., 2002). The propensity to invest was highest in agricultural processing (60%) which also has a high capital intensity followed by the fishing sector (50%). In the other manufacturing sectors less than half of the firms invest in a given year. However, the propensity to invest increases significantly with size. The large proportion of big enterprises and subsidiaries of multinational firms in the agricultural processing sector (which consists mainly of food and beverages) explains the higher percentage of firms that invest, as access to finance is less problematic for these firms. Most of the smaller firms reported that they finance investments from profits. From 2000 to 2002 the propensity to invest increased from 32% to 57% and this increase occurred across all size categories and all sectors except textiles and garments. However, to benefit from AGOA some new garment firms were established recently which 10
are not captured in the survey. For the existing textiles and garments enterprises AGOA seems to have little impact so far. Table 7: Propensity to Invest and Investment Rates (%) by Size and Sector, av. 2000-2002 size category (number of employees) Propensity to invest a less than 20 20 to 75 more than 75 Total horticultural 0.33 0.50 0.33 0.38 fish NA 0.75 0.44 0.50 other agricultural 0.00 0.30 1.00 0.29 agric. processing 0.45 0.44 0.71 0.60 textiles/garments 0.33 0.31 0.54 0.41 wood and furniture 0.33 0.44 0.40 0.38 aluminium 0.38 0.20 0.43 0.36 plastic and rubber 0.00 0.36 0.39 0.36 other manufacturing 0.25 0.57 0.43 0.41 handicrafts 0.06 0.14 NA 0.08 Total 0.26 0.39 0.52 0.41 Average investment rate b horticultural c 0.47 0.08 2.12 0.93 fish NA 0.20 0.00 0.07 other agricultural NA 0.05 0.08 0.07 agric. processing 0.19 0.11 0.10 0.11 textiles/garments 0.10 0.07 0.02 0.05 wood and furniture 0.07 0.06 0.44 0.16 aluminium 0.06 0.18 0.10 0.10 plastic and rubber NA 0.40 0.12 0.20 other manufacturing 0.16 0.13 0.00 0.08 handicrafts 0.28 NA NA 0.28 All 0.16 0.14 0.23 0.19 Notes: a - investment propensity means the likelihood that a firm in a specific category invests. b - amount invested divided by the total capital stock if a firm invests c - the high average investment rate for large firms in the horticultural sector is driven by one firm that started production on a low level in 1999 and expanded in 2000 The amount of investments a firm makes if it invests at all is 19 % of the total capital stock on average. This level is also comparable with the results of other studies (Rankin et al, 2002). Apart from the horticultural sector where the high investment rate is driven by one relatively newly established firm the investment ratio is highest for the handicraft producers. However, in this sector the capital stock is very low so that absolute investment is still rather low. Also in the plastic and rubber sector which is relatively capital intensive, investment rates are relatively high which is mainly driven by the growth of medium sized enterprises. Small enterprises have on average a higher investment rate than medium sized enterprises if they invest but they are less likely to invest at all. This reflects the problems of access to credit for small enterprises so they can only invest if they have accumulated enough profits. In addition small firms are constrained by the fact that investment in office equipment, machines or vehicles has to be made in one large sum as leasing is not very widespread. A number of small firms is also located in residential areas which also constraints their ability to expand. The average amount of new investment (in real terms) declined between 2000 and 2002. However, in 2003 it increased again considerably, above the 2000 level. The differences in investment behaviour between exporting and non-exporting enterprises in the sample are smaller than expected. However, more of the non-exporting enterprises also don t invest and more of the non-investing enterprises also don t export. Exporting enterprises are also more likely to invest (see Table 8). 11
Table 8: Export and investment categories, average 2000-2002 investment category Total exporting category no yes Total % of exporting category 57.9% 42.1% 100.0% % of investment category 38.0% 35.2% 36.8% % of exporting category 54.9% 45.1% 100.0% % of investment category 62.0% 64.8% 63.2% % of exporting category 56.0% 44.0% 100.0% % of investment category 100.0% 100.0% Note: The figures in the upper lines add up horizontally, whereas the figures in the lower lines add up vertically. For example the first figure in the first line means that 57.9% of those enterprises that don t export also don t invest. The figure in the second line first colum means that 38.0% of those enterprises that don t invest also don t export. no yes 6. Performance indicators for private enterprises Employment growth On average a decline in the number of employees can be observed. Between 2000 and 2002 the total number of employees in the sample firms declined by 4.3% (see Table 9). In 2003 the number started to increase again but average employment is still lower than in 2000. This reduction in employment is mainly driven by the bigger enterprises which tried to reduce costs by reducing their labour force. Partly this is driven by the general trend to outsource activities such as security and transport. Only some very successful bigger enterprises in the wood and furniture and plastic and rubber sector increased the number of employees over the study period. In contrast most of the small enterprises increased their number of employees except the horticultural sector. As the small enterprises are relatively young on average, many are still in the phase of getting the business started. However, this growth in employment might be partly offset by the closing down of other small enterprises which are not captured in the survey. Medium sized enterprises reduced their workforce mainly in the aluminum and handicraft sectors but increased it in the other sectors. Table 9: Change in employment by size and sector between 2000 and 2002 size category (number of employees) Sector less than 20 20 to 75 more than 75 All horticultural 0.0 0.0-55.4-48.2 fish NA 180.0-16.6-10.5 other agricultural 51.3 0.0-33.3-21.7 agric. processing 179.3 28.4-5.5-3.7 textiles/garments NA 42.3-16.8-10.9 wood and furniture 32.2 30.0 3.9 6.0 aluminium 81.5-36.2-6.1-5.2 plastic and rubber NA 75.6 3.3 4.8 other manufacturing 16.7 NA -4.3-4.1 handicrafts 34.5-28.6 NA 10.8 All 55.7 27.9-6.6-4.3 In sum the employment creation by small enterprises did not offset the loss by the big enterprises. However, as large firms are oversampled this might not be the case for the whole economy. In line with the increase in 2003 more than 50% of the firms were expecting their 12
labour force to increase, whereas only 12% expected a decrease. The percentage of small and medium firms that expected an increase in their workforce was higher than average and more large enterprises than on average expected a decrease in employment. The costs of labour are determined by a number of factors. The difference in wages and total labour costs for permanent workers are very high in Ghana due to a number of allowances (travel, food, housing) and benefits. On the other hand there are casual workers whose wages follow the minimum wage rate and who don t enjoy any benefits. Casual workers are not only employed in sectors with highly seasonal production but also to save costs. In addition it is quite difficult to lay off permanent workers. However, there are cases where casual employees had to be made permanent after they were employed for more than a year. Wages differ largely by size and sector. On average wages are three times higher in large enterprises than in small or medium enterprises. Partly this is driven by the fact that firms with some foreign ownership pay higher wages. Wages are also highest in the aluminum and agricultural processing sector. In the agricultural sectors wages are generally low but lowest wages are paid by handicraft producers. In general in the Northern Region wages are significantly lower than in the Greater Accra Region. Most of the companies (87%) give regular training to their employees as the education in the secondary schools and polytechnics is not practical enough for their purposes. The percentage of firms that give training to their workers increases with size (79% of small enterprises, 83% of medium sized enterprises and 95% of large enterprises). In the textiles and garments, wood and furniture, plastic and rubber and other manufacturing sectors all firms give some training to their workers. On the other hand less than 40% of all firms stated that they had difficulties with finding employees with the needed skills. For small enterprises this problem was only faced by 25 % of all firms. However in other agricultural production and plastic and rubber production this share was above 50%. However, a number of enterprises complained about the attitude of workers which is characterised by low motivation and frequent absenteeism especially of the permanent workers. A number of companies with production sites in different countries claimed that labour productivity in Ghana is only half of other developing countries, using the same equipment. Development of output Contrary to the developments in the number of workers, real output of the sample enterprises increased by 24.9% on average from 2000 to 2002 (see Table 10). However, huge differences occur across sectors and size categories. In the agricultural processing and plastic and rubber sectors, real output grew on average in firms of all sizes. In handicrafts production and the fisheries sector firms were also able to increase output. Most of the other sectors experienced a decline in output. On average for all sectors, output increased by only 15,6% for small enterprises whereas it increased by 17% for medium and 25% for large enterprises. This is also in contrast to the development of employment and indicates that for small enterprises labour productivity might have declined whereas it has increased for larger enterprises. This development of output per sector might be driven by a number of factors. As plastic and rubber products as well as beverages (that constitute a large part of agro-processing) face high transport costs, there is little competition from outside and as these sectors are relatively capital intensive, market entry of new domestic firms is also difficult. Furthermore output growth in the different sectors depends on the income elasticities of the products. There is some indication that higher incomes of cocoa producers have boosted demand for beverages and plastic and rubber products as products like beer, soft drinks, mattresses and cool boxes are luxury goods in the Ghanaian context. 13
Table 10: Real output growth by size and sector between 200 and 2002 size category (number of employees) Sector less than 20 20 to 75 more than 75 All horticultural -11,9-4,4-5,4-8,1 fish NA NA 64,7 64,7 other agricultural -1,7-28,2-46,7-42,8 agric. processing 31,2 13,2 37,1 37,0 textiles/garments NA -6,6-10,0-10,0 wood and furniture 17,8-10,9-23,1-22,3 aluminium 31,5-4,8-6,5-5,9 plastic and rubber NA 58,5 70,3 69,7 other manufacturing NA NA 4,2 4,1 handicrafts 56,0 NA NA 56,0 All 15,6 17,1 25,0 24,9 Financial performance indicators Besides firm growth that can be measured as increase in labour, output and export performance, a number of other indicators are used to measure firm performance. The rate of return (RoR) measures profits as a percentage of total capital. On average it has increased from 2000 to 2002 (see Table 11). This value is highest in the handicraft sector which operates with very low levels of fixed investment. For the plastics and rubber manufacturers and agricultural processing the RoR is also relatively high. In the fishing industry, other agricultural production and textiles and garments production the RoR is lowest and even negative for some years. Table 11: Rate of return by sector, 2000 to 2002 Sector 2000 2001 2002 horticultural -0.26 0.10 0.06 fish 0.02-0.02 0.35 other agricultural -1.03 0.17-0.64 agric. processing 0.11 0.20 0.22 textiles/garments 0.08-0.04-0.07 wood and furniture 1.32 0.55 0.02 aluminium 0.18 0.11 0.07 plastic and rubber -0.01 0.33 0.25 other manufacturing 0.27 0.42 0.07 handicrafts 6.22 1.06 1.40 All 0.10 0.14 0.14 Likewise the profit margin, which is measured as profits per value of output, increased from 2000 to 2002 (see Table 12). Again the handicraft sector is among the top performers together with agro-processing and plastic and rubber producers. For the agricultural sub-sectors, both RoR and profit margin have been very volatile. In the aluminium sector the relatively high RoR and profit margin have been declining significantly. In agro-processing, capital intensity, which is measured as capital per employee is lowest and here labour productivity, measured as value added per worker is highest. On the other end of the spectrum is handicrafts which has both the lowest capital intensity and the lowest labour productivity. Capital intensity increases on average with size according to expectations but labour productivity is lowest for the medium sized firms in the sample. 14
Table 12: Profit Margin by sector, 2000 to 2002 Sector 2000 2001 2002 horticultural -0.71 0.19 0.33 fish 0.02 0.00 0.29 other agricultural -0.69 0.22-1.06 agric. processing 0.10 0.18 0.24 textiles/garments 0.11-0.09-0.15 wood and furniture 0.18 0.08 0.01 aluminium 0.12 0.08 0.07 plastic and rubber 0.00 0.18 0.12 other manufacturing 0.11 0.17 0.05 handicrafts 0.23 0.20 0.21 All 0.09 0.14 0.16 A few preliminary observations can be made on what drives the differences in performance in the private sector in Ghana. 4 Firms that are more successful exporters in general use less imported inputs and are also less capital intensive, which includes less use of imported machinery. However, higher export orientation was not correlated with growth of employment, output or labour productivity. Firms that imported more of their inputs were able to increase output which reflects the fact that especially the rubber and plastic sector is highly dependent on imported inputs. These firms also have a relatively high labour productivity and capital intensity. Furthermore firms with some foreign ownership tend to import more as they have closer linkages to foreign suppliers, especially the multinationals. Surprisingly the investment ratio and the profit margin are negatively correlated, which is in contrast to the assumption that investment is mainly financed by profits. Firm productivity To measure total factor productivity, output has to be related to all inputs in the production process via production functions. 5 Besides the direct factors of production we included different firm characteristics as control variables in the regressions to test whether they have an influence on total factor productivity. A production function can be estimated either by regressing gross output on raw materials, indirect costs, employment and capital. This has the advantage that differences in efficiency of resource use can be measured. However, the use of raw materials is correlated with the capital stock and therefore it is difficult to determine the effect of capital on output. Therefore we also regress value added on employment and capital. The results of the regressions are presented in Table 13. In the first regression employment and capital are not significant because of the above mentioned problems. Also the age of the firm has no significant impact on performance. However, if a firm gives training to its workers output is increased by 1.11%. In the value added specification the coefficients on employment and capital are highly significant. A 1% increase in employment leads to a 0.39% increase in output, whereas a 1% increase in capital leads to a 0.32% increase in output. The positive effects of giving training to workers are also confirmed. This specification also suggests that older firms are slightly more productive, probably due to their experience. Other factors such as the share of foreign ownership and the export orientation of 4 5 The observations below are based on statistically significant correlations of two variables. As a Cobb-Douglas production function is used for the estimations, output and inputs are measured in logarithms. 15
a firm had no significant effect on productivity and were therefore omitted from the regressions. Table 13: Estimation of Production Functions log (real output) log (real value added) Constant 4.42 *** 10.56 *** log (raw materials) 0.17 *** log (indirect costs) 0.66 *** log (employment) 0.06 0.39 ** log (capital stock) 0.01 0.32 *** training to workers 1.11 *** 1.32 * age of firm 0.01 0.04 *** Adjusted R 2 0.948 0.784 No. of observations 152 146 Notes: * indicates significance at the 10% level, ** at the 5% level and *** at the 1% level The regression equation is of the following form: log (real output) = α 0 + α 1 log(raw materials) + α 2 Log(indirect costs) + α 3 log (employment) + α 4 log (capital stock) + α 5 dummy (training to workers) + α 6 age of firm. We also controlled for year, location and sector but the coefficients were not significant and are therefore not reported. The adjusted R 2 is a measure of how well the model fits the data. 7. Main Business Obstacles General business obstacles The perception of major business obstacles hasn t changed much over the past years. One major obstacle mentioned by almost all companies is the difficulty in accessing credit and especially the high interest rates. On average enterprises rate interest rates as the most problematic area (62% rate it as very or extremely problematic) and access to credit as second most problematic (52% rate it as very or extremely problematic) (see Figure 3). The limited access to credit is not only problematic for investments where long term capital from banks is virtually unavailable, especially for smaller firms. One reson for that is the non-availability of collateral due to problems with the registration of land titels. In addition it puts a constraint on the availability of working capital so many small firms have to rely on supplier credit or prefinancing by customers which restricts their sources of inputs and marketing possibilities. One reason for the low availability of credit is the high borrowing of the government which also leads to high interest rates. There is also a high government involvement in the banking sector. Of the 17 banks operating in Ghana, three are wholly state owned but in many others the state is directly or indirectly involved (IMF, 2003). On the other hand for bigger firms with less difficult access to credit, in general, it is not possible to give suppliers credit to retailers as enforcement of such contracts is very weak. Further obstacles for doing business in Ghana that were rated extremely or very problematic in our survey (in decreasing order) are depreciation of the Cedi, high inflation, tax regulations, high import tariffs and poor infrastructure (see Figure 3). Further obstacles known from similar business surveys include corruption and buerocratic red tape which increase the time needed to establish a business and the uncertainty for entrepreneurs. A major obstacle to exporting within the region is the lack of implementation of the free trade regulations in the ECOWAS states. Most of the countries in the West African Economic and Monetary Union (WAEMU or UEMOA) favour imports from WAEMU countries over Ghanaian products. Nigeria has banned imports of certain goods (e.g. textiles) without exemption for ECOWAS products. In addition transport costs are increased by a large number of road barriers. On the import side there are several ways of circumventing Ghanaian tariffs. 16
This is done through direct smuggling, under-invoicing of goods or declaration as transit goods that end up on the Ghanaian market. Although the government has proclaimed a Golden Age of Business and created a Ministry for Private Sector Development, many enterprises do not regard government policies as helpful and state that the business environment has not improved. Figure 3: Main business obstacles, 2003 interest rates access to credit depreciation inflation tax regulations high import tariffs insufficient infrastructure difficulty in aquiring land cost of domestic raw materials implementation of ECOWAS access to domestic raw materials industrial policies import/export procedures access to foreign currency low import tariffs shortage of skilled labour insufficient demand labour regulations 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% not little somewhat quite very extremely Note: The obstacles are ranked according to the percentage of firms that rated them extremely or very problematic. Delays of reforms in the energy and water sector The low quality of water and electricity supply causes severe problems for the manufacturing sector. Electricity is most often interrupted in Tema (average 8 days per month), followed by Accra (5 days per month). In Tamale the situation is somewhat better probably because relatively little energy is consumed there. Water is more often interrupted in Accra (3 days per month) than in Tema (2 days per month). In Tamale Industrial Area water is supplied only once per week so companies that use a lot of water have to install large water tanks or purchase water from water tankers. These shortages lead to frequent interruptions in production. Therefore around two thirds of the surveyed manufacturing enterprises (almost all of the larger enterprises) have a power generator and a water storage facility. It is estimated that in 1999 industry had an installed capacity of ca. 100 MW in generators of which 52 MW were installed in the mining sector alone. Power fluctuations also result in destroyed machinery, especially in the plastic sector and for modern electronic machinery. Whereas for electricity supply half of the enterprises saw an improvement, for water supply this share was only 45%. 17