©2017 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org Disclaimer This is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The manuscript of this paper has not been prepared in accordance with the procedures appropriate to formally edited texts. Rights and Permissions The material in this work is subject to copyright. Since The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Abstract The 2016 Malawi Investment Climate Assessment (ICA) evaluates the performance of domestic private sector firms in Malawi, based on the analysis of data from the 2014 Malawi Enterprise Survey. It identifies the barriers to increasing private investment and analyses productivity of enterprises in Malawi. This report also sets out prioritized policy recommendations aimed at addressing the constraints faced by enterprises. Authors: Efrem Chilima, William Mwanza, George Clarke and Richard Record. 2 GOVERNMENT FISCAL YEAR July 1 to June 30 CURRENCY EQUIVALENTS Currency Unit : Malawian Kwacha WEIGHTS AND MEASURES Metric System ABBREVIATIONS AND ACRONYMS ACB Anti-Corruption Bureau ATM Auto Teller Machine ATS Automated Transfer System CSD Central Securities Depository DFI Development Finance Institutions EGENCO Electricity Generation Company ESCOM Electricity Supply Corporation of Malawi ESW Economic Sector Work FDI Foreign Direct Investment GCI Global Competitiveness Index GDP Gross Domestic Product GOM Government of Malawi ICA Investment Climate Assessment IFMIS Integrated Financial Management Information System LP Labor Productivity MCCCI Malawi Confederation of Chambers of Commerce and Industry MFI Microfinance Institutions MSMEs Micro Small and Medium Enterprises MW Megawatts POS Point of Sale RBM Reserve Bank of Malawi SACCO Savings and Credit Cooperative Organization T&C Trade and Competitiveness TFP Total Factor Productivity 3 Vice President : Makhtar Diop Country Director : Bella Bird Country Manager : Laura Kullenberg Senior Director T&C : Anabel Gonzalez Director T&C : Klaus Tilmes Practice Manager T&C : Catherine Kadennyeka Masinde Task Team Leader : Brian Mtonya/Efrem Chilima Principal Authors : Efrem Chilima, William Mwanza, George Clarke, Richard Record 4 ACKNOWLEDGEMENTS The Malawi Investment Climate Assessment (ICA) 2016 was prepared by a World Bank team led by Efrem Chilima (Senior Private Sector Specialist), under the guidance of Catherine Kadennyeka Masinde (Practice Manager), Laura Kullenberg (Country Manager), Yutaka Yoshino (Program Leader), Steven Dimitriyev (Lead Specialist, T&C), and Brian Mtonya (Senior Private Sector Specialist). The co-authors were William Mwanza (Consultant), George Clarke (Consultant) and Richard Record (Senior Economist). The authors benefited greatly from advice from peer reviewers, namely: Julian Latimer Clarke, Moses Kibirige, Madalo Minofu, and Hardwick Tchale. The preparation of this document benefited from collaboration and support from Esther Lozo, Zeria Banda, Kehinde Vivian Waterman, Tillmann Von Carnap-Bornheim, Ellie Stylianou, Priscilla Kondoole and Mwayi Nampeya. Govati Nyirenda provided the photographs in the publication. The authors greatly benefited from interactions with stakeholders in Malawi including Government officials from the Ministry of Industry, Trade and Tourism. The team would like to thank the Enterprise Survey Unit for conducting the Enterprise Survey in Malawi. 5 Table of Contents Executive Summary .................................................................................................................................... 10 CHAPTER 1: Introduction ............................................................................................................................ 13 1.1 Background to the Economic Sector Work ................................................................................. 13 1.2 Investment Climate Assessment (ICA) ........................................................................................ 13 1.3 Enterprise Survey ........................................................................................................................ 14 1.4 Challenges With Using Enterprise Surveys ................................................................................. 14 CHAPTER 2: Context − Economy and Firm Competitiveness ....................................................................... 17 2.1 Low Income Agricultural Economy ............................................................................................. 17 2.2 Agriculture As The Main Sector .................................................................................................. 18 2.3 Financial Sector ........................................................................................................................... 18 2.4 Private Sector .............................................................................................................................. 21 2.5 Firm Competitiveness ................................................................................................................. 21 2.6 Infrastructure .............................................................................................................................. 23 CHAPTER 3: Enterprise Performance: How Does Firms Level Productivity in Malawi Compare with Other Countries and Across Different Categories? ............................................................................................... 25 3.1 Partial Productivity Measures: Definitions and Cross Country Comparison ............................... 26 3.2 Partial Productivity Measures for Different Categories of Firms According to Size, Location, Domestic or Export Oriented, Local or Foreign Owned and Whether or Not Credit Constrained .......... 27 3.3 Total Factor Productivity (TFP): Cross Country Analysis and Comparison with respect to Location in Malawi .................................................................................................................................. 30 3.4 Labor Costs and Productivity: Country Comparison and with respect to location and firm size31 3.5 Employment Growth .................................................................................................................. 34 3.6 Total Factor Productivity and Investment Climate ..................................................................... 34 3.7 Policy Implications of Productivity Measures ............................................................................. 35 CHAPTER 4: Enterprise Constraints: What do Firms in Malawi View as The Major Challenges to Enterprise Development and Growth? ........................................................................................................................ 38 4.1 Top three obstacles by Firm Size, Business Sector, Location and Gender Perspective............... 38 4.2 Comparison with 2008 Survey .................................................................................................... 40 4.3 Cross Country Comparison on Top Most Obstacles .................................................................... 41 4.4 Comparison with the World Bank Doing Business Survey Results ............................................. 41 4.5 Comparison with the MCCCI Business Climate Survey ............................................................... 42 CHAPTER 5: Looking at Top Constraints in Detail ....................................................................................... 44 5.1 Access to finance is a major constraint ....................................................................................... 45 5.2 Electricity outages ....................................................................................................................... 51 6 5.3 Corruption ............................................................................................................................ …...55 CHAPTER 6: Other risk measures ................................................................................................................ 60 6.1 Tax Rates and Administration ................................................................................................ …...60 6.2 Practices of the informal sector ............................................................................................ …...61 6.3 Access to land and Political Instability .................................................................................. …...62 6.4 Business licensing and permits ................................................................................................... 63 6.5 Crime, theft and disorder ........................................................................................................... 65 6.6 Transport and Logistics ............................................................................................................... 68 6.7 Water Shortages ......................................................................................................................... 69 CHAPTER 7: Conclusions for Policy on Investment Climate in order to improve Productivity ................... 70 BIBLIOGRAPHY ............................................................................................................................................ 74 Annexes Annex 1: Measuring Firm Performance Using Enterprise Survey Data .................................................. 77 Annex 2: Cross Country Comparison Using Enterprise Survey Data ...................................................... 80 Annex 3: Estimating Total Factor Productivity Using Enterprise Survey Data ....................................... 82 Annex 4: TFP and Investment Climate ................................................................................................. 90 Annex 5: What Is an Enterprise Survey................................................................................................. 93 Annex 6: Enterprise Survey in comparison to Doing Business methodologies ...................................... 94 TABLES Table 1: Basic Statistics for 2014 Malawi Enterprise Survey ...................................................................................... 15 Table 2: Banking Sector Market Share as at December 2014 ...................................................................................... 19 Table 3: Breakdown of Partial Productivity Data by Firm Type ...................................................................................27 Table 4: TFP and Improvement in the Investment Climate...........................................................................................35 Table 5: Three Biggest Obstacles to Establishments .....................................................................................................39 Table 6: Access to Finance, Electricity and Corruption ................................................................................................41 Table 7: Finance As a Major Obstacle ..........................................................................................................................45 Table 8: Firms with credit lines.....................................................................................................................................46 Table 9: Collateral Requirements ..................................................................................................................................47 Table 10: Working Capital Financed by Banks .............................................................................................................49 Table 11: Firms Using Banks to Finance Investment ....................................................................................................50 Table 12: Firms Identifying Electricity as Major Constraint .........................................................................................53 Table 13: Power Outages in a typical year in 2008 and 2014 ......................................................................................54 Table 14: Corruption and courts as Major Constraints ..................................................................................................57 Table 15: Firms Identifying Corruption as Major Constraint ........................................................................................58 7 Table 16: Corruption Levels – Country Comparison .................................................................................................... 59 Table 17: Tax Rates as a Major Constraint ................................................................................................................... 61 Table 18: Informal Sector Competition as Major Constraint ........................................................................................ 62 Table 19: Access to Land and Political Instability........................................................................................................ 63 Table 20: Business Licensing and Permit ..................................................................................................................... 64 Table 21: Losses Due To Theft, Crime, Vandalism...................................................................................................... 66 Table 22: Transport Challenges .................................................................................................................................... 68 Table 23: Water Insufficiencies .................................................................................................................................... 69 Table 24: Correlation Between Country Level Estimates of Technical Efficiency ...................................................... 84 Table 25: Breakdown of Partial Productivity Data by Firm Type................................................................................. 92 FIGURES Figure 1: Illustration of details of the survey Sample ................................................................................................... 15 Figure 2: GDP Growth Rates, Exchange Rates and Foreign Exchange Positions ........................................................ 17 Figure 3: Holdings of Credit By Private Sector in Malawi ........................................................................................... 20 Figure 4: Labor Productivity ........................................................................................................................................ 25 Figure 5: Capital Intensity ............................................................................................................................................ 25 Figure 6: Capital Productivity ...................................................................................................................................... 25 Figure 7: ToTal Factor Productivity ............................................................................................................................. 25 Figure 8: Large Firms are More LABOR Productive and Capital Intensive ................................................................ 28 Figure 9: Blantyre Firms are more Productive ............................................................................................................ 29 Figure 10: Labor Costs ................................................................................................................................................. 33 Figure 11: Unit Costs.................................................................................................................................................... 33 Figure 12: Employment Growth ................................................................................................................................... 33 Figure 13: Major Obstacles identified in percentage shares ......................................................................................... 38 Figure 14: Top Ten Biggest Obstacles to Enterprises in Malawi, 2008 and 2014 ........................................................ 40 Figure 15: Top Three Obstacles - Country Comparison ............................................................................................... 41 Figure 16: Top Obstacles According to Survey By MCCCI ........................................................................................ 43 Figure 17: Access to Finance as Major Constraint - Country Comparison .................................................................. 45 Figure 18: Firms Owning Bank credit line /Loan - Country Comparison .................................................................... 46 Figure 19: Loans Requiring Collateral -Country Comparison ...................................................................................... 47 Figure 20: Comparison of Value of collateral and NPLs – Malawi and Comparator Countries in year of enterprise survey .................................................................................................................................................................. 48 Figure 21: Percentage of Firms Using Banks to Finance Working Capital .................................................................. 49 Figure 22: Firms Using Banks to Finance Investment (%) ........................................................................................... 50 Figure 23: Investment Financing Options (proportion, %) ........................................................................................... 50 Figure 24: Electricity Outages And Associated Losses Across Countries .................................................................... 52 8 Figure 25: Firms Owning or sharing Generators (%) ....................................................................................................52 Figure 26: Number of Days to Obtain electricity connection ........................................................................................53 Figure 27: Firms Identifying Corruption as a Major Constraint (%) .............................................................................57 Figure 28: Bribery Rates ...............................................................................................................................................58 Figure 29: Top Ten Biggest Obstacles to Enterprises in Malawi, 2008 and 2014.........................................................60 Figure 30: Licensing, Taxation and Tax Administration...............................................................................................61 Figure 31: Competition from Informal Players .............................................................................................................62 Figure 32: Land and Political Instability as Obstacles - Country Comparison ..............................................................63 Figure 33: Firms Identifying Business Licensing and Permits as Major Constraints (%) .............................................64 Figure 34: Number of Days to get Licenses and Permits .............................................................................................65 Figure 35: Crime, Theft and Disorder - Country Comparison .....................................................................................66 Figure 36: Transport as a Major Constraint -Country Comparison ...............................................................................68 Figure 37: Water Insufficiencies in a Typical Month....................................................................................................69 Figure 38: Malawi Labor Productivity - Higher than would be expected given the level of development ...................80 Figure 39: Malawi Capital intensity - higher than would be considering the country's level of development ..............80 Figure 40: Malawi's Labor productivity and capital intensity are high but capital productivity is low .........................80 Figure 41: TFP IS Higher in Malawi Than Countries at Similar Level of Development ..............................................80 Figure 42: Despite High Labor Costs, Unit Cost in Malawi Remains Reasonable .......................................................81 Figure 43: Over the Past Three Years, Employment Growth was Slow Among Manufacturing Firms ........................81 Figure 44: Few Firms In Malawi Provide Workers With Formal Training ...................................................................81 9 EXECUTIVE SUMMARY The World Bank conducts investment climate assessments (ICAs) across the globe to analyze private sector performance and to determine principal constraints affecting the private sector in respective countries. The 2016 Malawi ICA focuses on supporting dialogue on investment climate by analyzing the challenges reported by private enterprises in the country. It is the second in a series, following the first one that was published in 2006. The 2016 Malawi ICA is based on the analysis of results from the World Bank enterprise survey conducted in 2014. Regression analysis on productivity levels of firms was conducted to present some insights into the competitiveness of manufacturing firms. The productivity analysis reveals that firms in Malawi have higher levels of labor productivity and capital intensity, but lower levels of capital productivity than their counterparts in neighboring countries. Unit labor costs are lower than in most comparator countries. The results are predominantly with respect to large firms, with medium firms in Malawi facing lower levels of labor productivity and capital intensity, but higher levels of labor costs than small and large firms. Apart from the foregoing measures of partial productivity, total factor productivity is lower than in most comparator countries. Turning to the analysis of the results of the enterprise survey themselves, the most frequently cited obstacles to growth and efficient performance of enterprises in Malawi are grouped in three main categories: finance, infrastructure, and business environment. The most significant constraint is access to finance, followed by unreliable electricity supply and frequent outages. The next most cited obstacles are the costs of dealing with corruption, particularly with regard to Government-to-business services and the burden of an unfavorable tax administration and rates. Other obstacles include unfair competition due to the unregulated practices of the informal sector, difficulty in accessing land, perceptions of political instability, excessive bureaucratic practices experienced when obtaining business licenses and permits, growing concerns over crime, theft and disorder, and high costs and poor transport infrastructure. The effects of limited access to finance are quite far-reaching, and felt more severely by Malawian owned local firms especially those that are rural-based, female headed, domestic market oriented, small scale and those engaging in the service industries. The high collateral requirements from commercial banks deter investment, especially by the small and the medium enterprises (representing the majority of enterprises in Malawi), resulting in heavy reliance on internal resources for financing investments, other than from the formal banking system. The interest rates are very high, sometimes running at around 40 per cent. Commercial banks offer only a narrow range of financial instruments. The country does not have long term financing schemes. As part of efforts to address this challenge, the Reserve Bank of Malawi is developing a leasing framework together with an examination of the applicability of a long term finance guarantee scheme. The deterioration in the reliability of electricity supply across Malawi forces many enterprises to sometimes convert to costlier diesel-powered generators. This increases their costs of production and operations, and worsens their overall performance. Per the survey, the manufacturing firms were the most affected by electricity outages. Another growing concern since 2008 is corruption, which was cited as the 3rd biggest obstacle in 2014, compared to 8th in 2008. Incidences of informal payments to public officials including tax officials, seem to have increased considerably. Perceived occurrences of bribery in form of “gifts” was noted when securing government contracts, obtaining business operating and import licenses, getting construction permits, meeting with tax officials, as well as getting connected to the electricity grid and water supply. 10 There are considerable challenges in dealing with the bureaucracy in government administration especially in obtaining business licenses and permits. Within categories of firms in the survey, medium-scale enterprises complained mostly about the exorbitant tax rates. The small-scale enterprises and the firms headed by female managers, mostly complained about the difficulty and costs of obtaining business licenses and permits. Growth of medium and small-scale enterprises was affected by practices in the unregulated informal sector. All the industry categories complained about the increasing difficulty to obtain rights to land1 for their business operations, and about increasing political instability. The former seems to have more severe impacts in Malawi than in some of the countries in the region. In contrast, business managers’ perceptions on transport challenges improved in 2014 compared to 2008. However, this is still worse than the average for the Sub-Saharan region. Moreover, the impact of transport and logistic challenges for Malawi is more pronounced because the country is landlocked. Similarly, the perception of crime, theft and disorder as a major constraint has lessened, but remains a concern to the private sector. Insufficient water supply is another infrastructure-related constraint to the growth of the manufacturing sector. Water disruptions occurred frequently in 2014 at 5.3 days a month on average (compared to 0.5 days in 2008). This was the worst performance amongst comparator countries in the region. The principal conclusion from the 2016 Malawi ICA is that policymakers face one distinct pathway if private sector will have to play its role as engine of growth; namely a radical reinforcement in the approach to private sector development, in which case, given all the challenges the country faces, the country’s pace of reform needs to be focused, be much more rapid, deep and consistent. This is necessary if Malawi is to enhance and maintain its competitiveness in a fast moving world. To do this, four priority steps are recommended: (1) Take advantage of Malawi’s higher levels of labor productivity and lower unit cost levels than neighboring countries to attract investment into labor-intensive industries. While doing this, labor productivity of medium firms needs to be improved, as does their use of capital. An increase in labor-intensive industries can create jobs that through increasing incomes, can positively reinforce capital utilization for domestic markets. Market access to regional and global markets can also enhance capital utilization as more products are produced for those markets. Further to this, total factor productivity needs to be improved inter alia through development of an innovation ecosystem and frameworks that facilitate high levels of technology diffusion; and research and development, education and skills development that promote agricultural production and feed into the labor needs of emerging manufacturing sub-sectors. (2) Ensure macroeconomic stability and provide financial infrastructure and instruments to reduce risks to lenders and the costs to borrowers with a view to effectively enhance access to finance by businesses. Based on the identified challenges, measures need to be taken to reduce interest rates and to prioritize the improvement of financial infrastructure to achieve broader access to finance. This should be complemented by measures to improve financial literacy and to develop financial products specially tailored towards the needs of small and medium-sized businesses, with attention to the disadvantaged identified across the different categories. This should also be accompanied by effective implementation of the warehouse receipt system, collateral registry, credit reference and long term financing infrastructure. 1 The World Bank supported the Ministry of Lands review 10 land laws which were enacted in 2016. The Bank has also supported the development of an integrated land information system and registry. 11 (3) Focus finite public sector resources in investment in sectors such as energy, water and transport that would provide the greatest contributions to growth of industries. Significant infrastructure improvements, especially for the availability of electricity (and water), should be undertaken if the productivity of firms is to be enhanced. This will require deeper efforts to improve the governance of utility suppliers and to encourage private investment to boost installed generation and transmission capacity. The generation capacity, including through alternative sources of energy, should be increased to ensure adequate supply of electricity to both rural and urban areas. Efforts to implement interconnectors with Malawi’s neighbors such as Mozambique and Zambia must also be reinforced to ensure that Malawi accesses energy from the Southern African Power Pool (SAPP). A dilapidated national grid network should be rehabilitated and expanded to uptake and efficiently transmit the increased generation of electricity that would result. (4) Address policy and regulatory challenges to increase predictability and transparency, and implement strong measures to curb corruption, especially in government to business services. Strong institutional measures are required to ensure that corruption and bribery are curbed. Regulatory processes including tax and licensing requirements should be made simpler, be online and automated and made more accessible to ensure these areas are not breeding grounds for corruption. This would ensure transparency and accountability, and help save the private sector from losing time and other resources. Improvement of IFMIS (the public financial management system) should be prioritized. The Anti-Corruption Bureau (ACB) and other governance institutions should be fully supported and financed. Implementation of reforms in land governance (new set of land laws) that have commenced should be effectively implemented to ensure predictability and ease in acquiring land by investors. Automation of land registration should be prioritized to ensure improvement on transparency. Government institutions mandated to provide land to investors, such as the Malawi Investment and Trade Centre, should be empowered to undertake their function effectively. Government should also ensure a more predictable and transparent investment climate for commercial agriculture and agribusiness, and specifically provide a legal framework that allows for clearly defined and transparent procedures for the regulation of commodity imports and exports. The review process of the Control of Goods Act should be expedited and Special Crops Act should be reviewed or repealed to remove unpredictability in this area. 12 CHAPTER 1: INTRODUCTION 1.1 BACKGROUND TO THE ECONOMIC SECTOR WORK The Malawi Growth and Competitiveness Economic Sector Work (ESW) is an analytical piece of work by the Trade and Competitiveness (T&C) Global Practice of the World Bank Group, designed in 2015 to complement the preparatory work for a lending operation - the Shire Valley Transformation Project (SVTP) - and to promote dialogue around the SVTP and Government’s overall strategy on agriculture commercialization. The SVTP is a World Bank project aimed at supporting the Government to develop a gravity-fed irrigation scheme covering around 42,000 hectares to the west of the Shire River in Chikwawa and Nsanje districts. The Bank is also financing another operation, the Malawi Agricultural Commercialization (AGCOM) project, which is supporting farm level productivity, with significant efforts to build productive alliances as an agglomeration strategy aimed at improving farmer access to inputs, technology, and markets. The Bank is also financing the Malawi Resilience Development Policy Financing (DPO) to address policy and regulatory reforms impeding the effective functioning of agricultural markets in order to achieve structural transformation. An analytical work was designed: (a) to identify private sector investment opportunities and spill overs from the SVTP in the Shire Valley area; and, (b) to identify key investment climate and trade policy/logistics constraints as well as necessary policy action reforms that need to be taken for the positive spill overs of the SVTP to result in increased investment. The 2016 Malawi ICA aims to support dialogue on the investment climate in the country by analyzing challenges reported by the private sector. This analyzes firm level constraints based on a World Bank Enterprise Survey that was conducted from April 2014 to February 2015. The survey was conducted in the form of interviews with business owners and top managers from 523 formal private firms, categorized into small, medium and large firms in the manufacturing, service and retail sectors. The survey covered six districts: Blantyre, Kasungu, Lilongwe, Mangochi, Mzimba, and Zomba.2 1.2 INVESTMENT CLIMATE ASSESSMENTS (ICAS) ICAs provide a standard for the measurement and comparison of investment climate conditions in a country, creating the platform for the identification of the features of the investment climate that matter most. ICAs track changes in the investment climate and make international comparison between countries regarding firm level characteristics and constraints. This kind of analysis provides insights to policy makers and development specialists to effectively formulate policies and propose steps that can be implemented to improve the business climate and increase investment. The Malawi ICA 2016 is the second report in series. The first was done in 2006 based on the 2003 World Bank Enterprise Survey. This report is based on analysis of the enterprise data which was collected and processed in 2014 and part of 2015. The Malawi Confederation of Chambers of Commerce and Industry conducted its own surveys in 2015 and 2016, and the results mirror the findings of the Malawi ICA 2016, indicating that the findings are still prevalent currently. The assessments drew comparisons between selected countries in the region3 and compared its findings with similar assessment conducted in 2008. It also analyzed some of the risks across geographical regions in Malawi, different sizes of enterprises, and analyzed productivity measures. 3 Thefollowing comparative countries in the Region were used. Note that the latest enterprise surveys in these countries were conducted at different times as indicated in brackets: Zambia (2013), Mozambique (2007), South Africa (2007), Zimbabwe (2016), Tanzania (2013), and Kenya (2013). 13 1.3 ENTERPRISE SURVEY A World Bank Enterprise Survey (WBES) broadly explores challenges faced by firms with respect to the most significant obstacles to the daily operations of their businesses, as well as the degree to which each element poses as an obstacle. First, the survey questionnaire requests respondents to indicate which element, from a list, represents the biggest obstacle faced by its establishment. The elements are ranked according to the percentage of firms that selected each constraint. The second question is on the degree to which each element of the business environment is an obstacle to the current operations of the establishment, whereby firms provide a rating in a range of 0 for ‘no obstacle’ to 4 for ‘very severe obstacle’. More details are provided in Annex 5. The Malawi 2014 enterprise survey is the third to be undertaken, building from previous assessments in 2008 and 20034. In 2014, the survey covered 523 business establishments registered as privately-owned with sampling stratified by sector, size, and gender of owner. The survey involved business owners and top managers who were interviewed between April 2014 and February 2015. The survey sample included panel interviews from those conducted in the previous round, so that the performances and constraints are compared over time. Table 1 and Figure 1 below provide some descriptive statistics of the 2014 Malawi Enterprise Survey. 1.4 CHALLENGES WITH USING ENTERPRISE SURVEYS ICAs and analyses from Enterprise Surveys have proven to be an important method of assessing the investment climate of countries. Despite concerns that subjective views and ranking constraints based on perception-based measures pose analytical challenges, this approach has been seen to provide useful insights. A strong assumption is that managers and owners know more about the constraints and obstacles facing their businesses hence their views are significantly representative of the actual situation (Clarke, 2010). For the ICA, regression analysis was conducted to determine labor, capital and total factor productivity levels in Malawi. This analysis was also conducted on the data from the Enterprise Survey, to determine the extent to which the identified constraints affect total factor productivity, and hence long term growth in Malawi. 4 The2003 survey covered 306 firms, and the 2008 survey covered 160 firms (World Bank, 2006) (World Bank, 2006, and World Bank, 2017c)). 14 Table 1: Basic Statistics for 2014 Malawi Enterprise Survey5 ______________________________________________________________________________________ Total Number of Survey Respondents 523 Small Enterprises (5-19 employees) 291 Medium Enterprises (20-99) 148 Large scale Enterprises (100+) 84 Location of Firms Blantyre 256 Lilongwe 183 Zomba 36 Mzimba 30 Mangochi 14 Kasungu 4 Sectors Manufacturing 197 Retail 118 Other 208 Source: World Bank Enterprise Survey 2014 FIGURE 1: ILLUSTRATION OF DETAILS OF THE SURVEY SAMPLE SURVEY SAMPLE BY LOCATION FIRM SIZE Kasungu Mangochi Mzimba Largescal Blantyre e Kasungu Zomba Smallscale Mangochi Medium Mzimba Zomba Lilongwe Lilongwe Blantyre BUSINESS SECTOR GENDER Manufacturing percentage of firms with a female top manager Other percentage of firms with a male top Retail manager Source: Authors’ calculation based on data from WBES 5 Itwill be noted that the districts included in the survey are based on the number of firms operating in them. There is quite some activity in other districts such as Chikwawa, Thyolo and Mulanje where there is agro-processing of sugar and tea, respectively. The analysis may have been helped by insights from firms in these districts. However, these were not in the scope of the Enterprise Survey. 15 16 Tea plantation in Thyolo CHAPTER 2: CONTEXT - ECONOMY AND FIRM COMPETITIVENESS The 2014 Malawi Enterprise Survey was conducted between April 2014 and February 2015. This was also a period when the country faced significant business challenges and an economic slowdown. The manufacturing sector was especially affected by insufficient electricity supply and a significantly depreciated currency. The manufacturing base was, and remains, small, dominated by informal Small and Medium Enterprises (SMEs). A few large firms operate in the country and only a few of them produce for export. The enterprises face stiff competition on the international markets because of high costs of imported raw materials, worsened by the landlocked6 position of the country. 2.1 LOW INCOME AGRICULTURAL ECONOMY Malawi is categorized as a low-income economy. The country’s Gross Domestic Product (GDP) per capita (current US$) was estimated at US$372.4 7 in 2015. About half of the population lives below the national poverty line and a quarter is estimated to live in extreme poverty. The poverty levels declined, but only slightly, from 52.4 percent to 50.7 percent between 2005 and 2011, a period when GDP grew strongly (on average by 6.5 per cent). This points to the worsening of the income distribution, not in favor of the majority of the population in Malawi living in the rural areas. Those living in rural areas, have their economic output revolving around a narrow resource base of primary agricultural commodities. Due to a combination of weather related shocks, macroeconomic instability and other economic challenges, the country’s GDP growth has not been steady over recent years (see illustration in Figure 2). FIGURE 2: GDP GROWTH RATES, EXCHANGE RATES AND FOREIGN EXCHANGE POSITIONS 7.0 5.0 GDP Growth Rates 4.0 3.0 2.0 1.0 Aug-16 Nov-13 May-14 May-15 May-16 Aug-14 Aug-15 Nov-14 Nov-15 Feb-14 Feb-15 Feb-16 Source: World Bank (2017a) Malawi experiences consistent macroeconomic shocks. Overall, the country’s foreign exchange market has been under pressure following the persistent depreciation of the Kwacha (from trading at MK405.2 to the US dollar in 2013 down to MK730.08 in November, 2016 – and at MK734.7 as at April, 2017). GDP growth has slowed in recent years from 5.7 percent in 2014 to 2.8 percent and 2.5 percent in 2015 and 2016, respectively. The manufacturing sector declined by 1.1 percent in 2012. In 2013/14, the country experienced lapses in public financial management with the systematic embezzlement of public funds which was dubbed “cashgate” leading to donors withholding budgetary support. The country also experienced high inflation rates, fuel shortages and frequent electricity load shedding. 6 Malawi shares borders with Zambia to the West, Tanzania to the North, and Mozambique to the South and East 7 Or US$493.7 in constant 2010 US$ prices (World Bank, WDI) 8 http://www.rbm.mw/Statistics/MajorRates 17 Malawi’s rate of inflation has been very high but decreased rapidly in 2017 due to good harvests, a stable exchange rate and efforts in implementing fiscal discipline. Inflation, which was at 34.6 percent in December, 2012, peaked at 37.9 percent in February 2013, before dropping to 25.9 percent in January 2014. The rate had consistently been above 20 percent since 2012, but started showing signs of declining in 2016/17. As a result of the consistently high inflation rates, interest rates were above 40 per cent at the time that the 2014 Enterprise survey was conducted. As at May 2017, the inflation rate had dropped to 12.3 percent. Malawi’s lending rates and interest spreads are very high. Since 2012/13, the nominal lending rate and interest rate spreads (gap between lending rates and deposit rates) have been widening. A number of key variables have contributed to the spread and the persistently high lending rates, including: borrower credit risk, high operating costs for banks and cost of funding as banks bid their large deposit base on high yielding treasury bills. Malawian banks have high exposure to the public sector which not only crowds out the private sector through heavy borrowing, but also exerts upward pressure on the interest rate (through high treasury bill yields). The country had neither a well-functioning credit reference nor collateral system for some time. In 2016, a collateral registry was established, and the credit reference system was strengthened, though these had not yet been widely used to have an impact on interest rates. The pre-2016 collateral system placed foreign owned firms at an advantage as they had larger immovable asset bases and were perceived less risky. 2.2 AGRICULTURE AS THE MAIN SECTOR Malawi’s economy is predominantly dependent on agriculture . The sector is dualistic in nature. It comprises the smallholder subsector which is dominated by maize production, and the large-scale sector focusing on the commercial agricultural cash crops such as tobacco, tea, and sugarcane. The agriculture sector accounts for 30 percent of GDP, generates over 80 percent of export earnings, and accounts for over 64 percent of the country’s workforce9. Several challenges make it difficult for the manufacturing sector to benefit fully from agriculture. These include: unpredictable agriculture output levels and poor quality and standards, predominance of maize staple in the sector (lack of diversification), and impediments to the manufacturing sector (details of these challenges will be discussed in subsequent chapters). Exports are predominantly primary agricultural, including tobacco, tea and sugar. The overall trend of external trade balance points to rapid growth of imports as compared to exports. The country’s imports are dominated by petroleum products, fertilizers, and medicine. Tobacco dominates Malawi’s export basket accounting for 60 percent in 2016.10 The country’s main export markets are in Europe, Asia, South Africa, and more recently Zimbabwe and Mozambique. 2.3 FINANCIAL SECTOR Malawi’s financial sector is small, shallow and commercial bank-dominated. When the enterprise survey was conducted in 2014, the banking sector comprised 11 banks, one discount house and one leasing company. Two banks, namely Malawi Savings Bank and Inde Bank, have since been taken over by FDH Bank and National Bank respectively, and a new bank has been established. The rest of the banking sector comprises two discount houses, one leasing company, four development finance institutions (DFIs), 22 microfinance institutions (MFIs), 41 savings and credit cooperative organizations (SACCOs), and a nascent capital market. There is also a small life and general insurance subsector. 9 Government of Malawi, 2013, Malawi Labor Force Survey, National Statistics Office 10 ITC Trade Map. Available online at: http://www.trademap.org 18 There are a number of challenges affecting the ability to borrow from financial institutions. The judicial system does not seem able to provide full protection to the banks as it appears to be too easy to obtain court injunctions stopping the financial institutions from enforcing loan recovery procedures in cases of default, which makes financial institutions tighten their lending policies, thereby making it difficult for average firms to qualify for credit.11 Loans in Malawi are sometimes construed to be free by borrowers especially in cases where Government or development partners are involved or are deemed to be providing some form of subsidy. Heavy default rates are also a result of unfairly high and non-transparent borrowing costs, with some institutions charging 120 per cent interest rates per annum. There is currently no robust system of disclosure of financial products by financial institutions. Banking Sector The country's major banks, National Bank and Standard Bank, account for over 50 percent of assets, deposits and capitalization, and command the majority of loan portfolio . See table 2 for details. TABLE 2: BANKING SECTOR MARKET SHARE AS AT DECEMBER 2014 Assets Loans Deposits Capital Name of Bank MK’ billion MK’ billion MK billion MK billion billion Share billion Share billion Share billion Share National Bank of Malawi (NBM) 206.3 26.3% 80.7 25.7% 136.3 25.1% 41.7 29.4% Standard Bank Malawi (STD Bank) 191.8 24.4% 57.6 18.3% 140.1 25.8% 37.0 26.1% First Merchant Bank (FMB) 81.4 10.4% 31.7 10.1% 46.2 8.5% 22.7 16.0% NBS Bank Limited (NBS) 75.3 9.6% 41.2 13.1% 49.1 9.0% 11.7 8.3% Ecobank Malawi Limited (Ecobank) 50.3 6.4% 22.8 7.3% 34.0 6.3% 5.7 4.0% Malawi Savings Bank (MSB) 49.8 6.3% 26.5 8.4% 40.9 7.5% 4.8 3.4% First Discount House (FDH) 49.5 6.3% 21.2 6.7% 39.1 7.2% 7.4 5.2% Investment and Development Bank 23.4 3.0% 13.3 4.2% 17.3 3.2% 1.7 1.2% (INDE Bank) Opportunity Investment Bank of 21.5 2.7% 6.5 2.1% 12.1 2.2% 3.5 2.5% Malawi (OIBM) Continental Discount House (CDH 19.1 2.4% 8.0 2.5% 14.6 2.7% 3.0 2.1% Bank) NED Bank Malawi Limited (NED 16.5 2.1% 4.8 1.5% 12.8 2.4% 2.4 1.7% Bank) Total 784.9 100% 314.3 100% 542.5 100% 141.6 100% Source: Reserve Bank of Malawi (2014a) Bank lending rates and credit risks to the financial system continue to be high in Malawi. Monetary conditions remained tight with the bank rate at 25 percent for most of 2013. The rate increased to 27 percent in November 2015. Commercial banks’ lending rates were as high as 42 per cent by December, 2016 12. Credit risks to the financial system increased as non-performing assets in the banking system (as a ratio of gross loans and leases) rose to 15.7 percent in March 2014 from 13.6 percent in September 2013. The outturn was mostly due to the high inflation rates that prevailed in the economy. Inflation has been on a downward trend from the second half of 2016, and this has led to the Reserve Bank of Malawi cutting the bank rate from 27 per cent to 24 percent in November, 2016 and 22 percent in March, 2017. 11 This is a challenge that possibly requires stakeholder consultations between the Judiciary, the Executive and all relevant stakeholders in the public and private sector. 12 http://www.rbm.mw/Statistics/BankRates 19 Credit to the private sector continued to increase in 2013/14 in favor of the corporate sector . The corporate sector (commercial and industrial) was the largest holder followed by credit to individuals as at March 2014. Credit to the agriculture sector was the lowest as at March 2014. FIGURE 3: HOLDINGS OF CREDIT BY PRIVATE SECTOR IN MALAWI Source: Authors’ calculations based on Reserve Bank of Malawi June 2014 Financial Stability Report Insurance and Pension Insurance and pension sectors are growing but are not fully developed. Total insurance assets – both general and life insurance - reached 9.7 percent of GDP in 2014, from 6.9 percent in 2010. Meanwhile, following the adoption of the 2011 Pension Act, the total assets of the pension sector grew to 9.6 percent of GDP in 2014 from 5.7 percent in 2010. Assets stood at MK365 billion in September 2016.13 E-Payment Innovations There were some improvements in financial sector innovation. In February 2015, the Malawi National Switch system was launched to facilitate inter-operability of ATMs and point of sale (POS) terminals in the country. This was the third major national payments infrastructure improvement to be implemented in the country following the Automated Transfer System (ATS) and Central Securities Depository (CSD) which went live in December, 2014. Since the introduction of non-bank led mobile payments in Malawi, the number of active subscribers to mobile payments has been growing. In terms of service distribution channels, there was notable growth of mobile money agents, contributing considerably to financial inclusion. Internet banking maintained an increasing trend and subscriber base for bank-led mobile payment schemes continue to register increases reaching 520,959 by November, 2016. The volume of transactions through this channel has significantly expanded. Financial Inclusion Financial inclusion improved in the period between 2008 and 2014 . According to the 2008 and 2014 Malawi Finscope Survey report, the banked population in Malawi increased from 18.9 percent in 2008 to 33 percent in 2014 (by about 2,390,605). A number of initiatives were instituted to develop institutions that 13 https://www.rbm.mw/FinancialStability/FinancialStabilityReports/ 20 would pursue the access to finance agenda. An example is the Export Development Fund, a development financial institution established in 2012 whose mandate is to increase the productive potential of the country through provision of finance, equity participation, performance bonds or guarantees, and advisory service for the set-up, expansion and modernization of viable enterprises in the medium and large scale enterprises sector. Beyond such kind of new initiatives, there has been a growth in the “village banking”, “mobile banking” services and other new similar innovations with outreach to even the small scale enterprises and the people living in rural remote areas of the country. 2.4 PRIVATE SECTOR The Private Sector is characterized by a few large private firms and a large number of informal small and micro enterprises. Large firms are often engaged in agriculture or agro-processing. The micro and small enterprises are primarily informal and operate in the services sector, mainly in retail. The 2012 Malawi FinScope MSME Survey estimated that there are almost a million MSMEs in the country, employing over a million people in total, with their total income equivalent to a sum greater than 30 percent of the annual GDP in 2011. Private Sector in Malawi is characterized by a “missing middle”. This absence of a larger group of middle-sized firms points to the difficulties in the business environment, which constrains the growth of small firms. Along with the difficult business environment, the risks that are brought about by persistent instability in the macro-economy and regulatory deficiencies are more navigable by larger firms due to their broad networks and larger financial resources. Private Sector development is undermined by a number of significant challenges. Enterprises operate under a challenging business environment such as macroeconomic instability, corruption (including on land acquisition), inconsistent policies, and access to good markets. Other challenges are infrastructural such as poor road networks, inadequate provision of electricity and water which further undermine private sector development. Interest rates in the informal sector can go as high as 60 percent while prime lending rates are still above 30 percent per annum making it prohibitive for the private sector to borrow, thus restricting investment. 2.5 FIRM COMPETITIVENESS While there have been some improvements recently according to the Doing Business (DB) reports, the country’s ranking remains low . The DB ranking worsened consistently for five years reaching its trough of 164 in 2014. According to DB17 Report, Malawi ranked low, mostly ranked outside the top 100 economies against all 10 indicators, including those affecting competitiveness: Starting a Business (150/190), Trading across Borders (118/190), Paying Taxes (102/190), and Getting Electricity (169/190). Overall, the country’s DB ranking has recently shown some improvement from 144 th in 2015 to 141st in 2016, and 133rd in 201714. Malawi fares poorly on the competitiveness indicators and rankings. The Global Competitiveness Index (GCI) placed Malawi at 135 out of 140 countries in 2015/16. The GCI attempts to quantify the impact of a number of key factors which contribute to creating conditions for competitiveness, with particular emphasis on the macroeconomic environment, the quality of the country’s institutions, and the state of the country’s technology and supporting infrastructure. Unbundling the index reveals low scores by Malawi on basic requirements for business competitiveness, including the macroeconomic environment and other efficiency enhancers such as education standards, market size and efficiency, financial markets, innovation and technological readiness. 14 http://www.doingbusiness.org/data/exploreeconomies/malawi/ 21 Power lines in Kanengo, Lilongwe 2.6 INFRASTRUCTURE Electricity In 2014, Malawi enterprises experienced one of the worst electricity crisis in years, characterized by a widespread national load-shedding program. At that point, the country had a total installed hydropower capacity of 351MW, generated from the Shire River. The peak demand was estimated to exceed 400 MW, and the deficit caused chronic shortage of power and increased frequency of unscheduled outages. The challenge of prolonged load-shedding continued through 2015 and 2016 due to low water levels in the Shire River because of severe droughts experienced in those years – making evident the continued vulnerability of Malawi’s energy security to weather shocks. To date, only nine percent of the population has access to electricity. For the 80 percent of the population who live in rural areas, the access rate is limited to only about four percent. Until recently, the Electricity Supply Corporation of Malawi (ESCOM), was the only Government-owned electricity utility that generated, transmitted and distributed power. By 2015, the customer base was about 312,857 (from about 200,000 in 2014), while the installed generation capacity remained at 351MW, against a demand potential around 440 MW. The energy sector has faced numerous other challenges, apart from limited investment in generation and reliance on one river as a source15, including inadequate transmission and distribution networks and relatively low tariffs that affect the economic viability of other sources of energy. These tariffs together with uncertainties associated with the regulation and governance of the energy and electricity sector – including limited autonomy of utility operations and the energy regulator from Government - have been limiting factors for reinvestment in the network and undertaking new investments in the sector. A number of reforms have been implemented in the energy sector. The power system in Malawi until recently has been vertically integrated with ESCOM being the only utility in charge of generation, transmission, distribution and sale of electricity. The Electricity (Amendment) Act was passed in 2016 to allow for the involvement of the private sector in generation. The new law also created a single buyer and a system and market operations segment in the power sector. This resulted in the unbundling of ESCOM into two segments i.e. Electricity Generation Company (EGENCO) to handle generation and ESCOM to handle transmission and distribution. Water Malawi enterprises face challenges in the provision of water. Per capita water availability has been declining at a rapid rate due to population growth which comes with increased demand. Further, water resources in Malawi are highly variable between wet and dry seasons and from year to year, and the country’s stock of water storage infrastructure is one of the lowest in the region. There is inequitable water supply coverage especially in rural areas, as well as non- functionality of water points. In the cities and towns, water supply is characterized by intermittent and unreliable supply. Some of the key issues and challenges in the sector include: (i) increasing climatic and hydrological variability and limited resilience to floods and droughts; (ii) limited stock of water storage and irrigation infrastructure; (iii) degradation of watersheds leading to increased soil erosion and sedimentation; (iv) inefficient water utilities that lack autonomy and accountability; and (v) limited maintenance of distribution infrastructure resulting in significantly high leakage rates. Transport 15 Malawi is still not connected to the Southern Africa Power Pool (SAPP), therefore not yet able to trade electricity through the SAPP. 23 Being landlocked, the transport sector plays a key role in the country- and firm-level competitiveness. Producers in Malawi face high trade and transport costs in sourcing inputs and in delivering outputs to domestic, regional, and international markets. Road transport, which is comparatively more expensive than other modes of transport such as rail and water, accounts for 70 percent of all freight, and 99 percent of all passengers. The overall quality of the road network, particularly at the secondary and tertiary level, and connectivity to some rural areas is poor, further increasing challenges of accessibility to markets. There have been efforts to develop other modes of transport in recent years. For example, the rail route from Tete province in Mozambique, through Malawi, to Nacala port has been rehabilitated. The full potential benefits for Malawi in terms of reduced import and export costs by utilization of this railway line is yet to be realized in part due to the absence of adequate port handling infrastructure in key production areas connected to the railway such as Lilongwe and Blantyre. The southern-most leg of the rail network as well as the connection to the central region of the country and to Zambia is not functional. 24 CHAPTER 3: ENTERPRISE PERFORMANCE: HOW DOES FIRM LEVEL PRODUCTIVITY IN MALAWI COMPARE WITH OTHER COUNTRIES AND ACROSS DIFFERENT CATEGORIES? This chapter looks at measures of firm-level productivity in Malawi, starting with comparisons on partial productivity, labor productivity, capital intensity, capital productivity and total factor productivity. The analysis focuses only on manufacturing firms interviewed in the Enterprise Survey (i.e. 197 out of 523 firms surveyed). This is the case in order to draw more accurate findings on the factors of production and to also draw more accurate comparisons of firms across the country and in comparator countries (details on this and the calculation of the productivity measures are provided in Annexes 1,2 and 3). Regression analysis was used to determine productivity levels in Malawi with respect to labor, capital and total factor productivity, as well as labor costs and growth in employment; and the extent to which the investment climate constraints identified in the enterprise survey affect total factor productivity. $25.0 $23.7 $12.0 $10.6$10.9 value added '000 US$ Value added '000 US$ $20.0 $15.2 $10.0 $15.0 $12.4 $8.0 $6.0 $4.3 $10.0 $4.5 $5.1 $6.7 $2.8 $3.0 $5.0 $2.1 $3.5 $4.0 $1.9 $1.3 $2.2 $2.0 Mozambique Tanzania Zambia Malawi India Kenya Zimbabwe South Africa Mozambique Tanzania Zambia Malawi India Kenya South Africa Zimbabwe 250% 200%196% 400% 344% 161%152%148% Median TE relative Capitla Productivity 200% 300% 240% 150% 115%115% 185% 200% 116%100%132% to Zambia 100% 57% 77% 100% (%) 0% 0% South Africa Kenya Zambia Tanzania Zimbabwe Mozambiq… Malawi Mozambique Tanzania Zambia Malawi India Kenya Zimbabwe South Africa Source: Authors’ calculation based on data from World Bank Enterprise Surveys (WBES) Note: The country estimates are from the LAD regressions. Notes: 1. Partial productivity measures are measured in 2010 US$. See annexes, 1,2,3 for details. All data points are for the median firm on each measure of performance. 2. For Figure 9.4, Country estimates are from the LAD regression 3. Comparisons made were for median firms in terms of the different measures of performance. Medians are preferred to means because means can be affected greatly by outliers—possibly due to misreported or miscoded information. 25 3.1 PARTIAL PRODUCTIVITY MEASURES: DEFINITIONS AND CROSS COUNTRY COMPARISON Labor Productivity Labor productivity is defined as the rate of output or value added per worker (or a group of workers) per unit of time. It is calculated by subtracting raw materials and intermediate input from sales and dividing by number of workers, and tends to be higher where firms substitute capital for labor. Although in theory this is only a partial measure, and does not take into account capital contribution, it does provide useful information and is easy to calculate. Labor productivity is considered higher for firms that produce more output with less raw materials and fewer workers (or where firms substitute more capital for labor). This reflects better worker skills, management ability, and use of technology. Figure 4 compares labor productivity for Malawi and other countries. Labor productivity tends to be higher in countries with higher per capita income. Analysis shows that Malawi has higher labor productivity in comparison to some of the countries in the region. The country’s labor productivity is higher than neighboring countries - Mozambique, Tanzania and Zambia, but it lags behind the levels of other countries such as South Africa, Zimbabwe and Kenya. At $5,100, Malawi’s labor productivity is higher than would be expected given the country’s low level of per capita income. One of the weaknesses of labor productivity as a measure of firm productivity is that it does not take into account the use of capital, i.e. equipment and machinery. Capital Intensity Capital intensity is a measure of the relative use of capital, compared to other factors such as labor, in a production process. It is defined as the amount of fixed or real capital present in relation to other factors of production. Sometimes capital intensity is also defined as a measure of a firm's efficiency in deployment of its assets, computed as a ratio of the total value of assets to sales revenue. This analysis considers capital intensity as capital to labor ratio, measured as the value of the firms’ machinery and equipment divided by number of workers. Median firms in Malawi appear relatively more capital intensive than some of the other countries in the region. According to this analysis, firms in Malawi are relatively more capital intensive, in comparison to similar firms in neighboring countries (Tanzania, Mozambique and Zambia) (See Figure 5 and detailed explanations in Annex 2). For Malawi’s case, deeper analysis shows that only large firms appear to be particularly capital intensive, over four times the capital intensity of SMEs. Just like the case of labor productivity, capital intensity is higher in Malawi than other countries at similar levels of development. Comparing Malawi with a broader group of countries at similar levels of development, the median firm in Malawi appears to be more capital intensive than the median firms in the other countries (at similar levels of development). Capital Productivity From the analysis above, both capital intensity and labor productivity are relatively high in Malawi, but it would be necessary to look at the value added of machinery and equipment. Capital productivity measures the ratio of value added to the value of machinery and equipment. It is the degree to which physical capital is used to provide goods and services. Capital productivity is high when firms produce more output with only a small amount of machinery and equipment. Analysis shows that capital productivity is low in Malawi. For the median firm in Malawi, capital productivity is about 115 percent, which is lower than in any of the comparator countries (see Figure 7). 26 Although both capital intensity and labor productivity are relatively high, capital productivity is not, implying that Malawi’s firms produce less output for the available machinery and equipment, that firms use capital inefficiently, or that firms are highly labor intensive. Enterprises in Malawi seem to fall under these explanations. This points to the need to improve usage of more efficient technology, as well as more technical skilled labor. 3.2 PARTIAL PRODUCTIVITY MEASURES FOR DIFFERENT CATEGORIES OF FIRMS ACCORDING TO SIZE, LOCATION, DOMESTIC OR EXPORT ORIENTED, LOCAL OR FOREIGN OWNED AND WHETHER OR NOT CREDIT CONSTRAINED This section presents a further breakdown of partial productivity measures for various groups of firms with sufficient data including firm size, location, whether the firm is an exporter or domestically or foreign owned and whether the firm is credit concerned. It was noted earlier on that the sample of manufacturing firms that actually provided performance data in Malawi is small (197 out of 523 firms surveyed). As a result, only a limited number of breakdowns will be provided.16 The labor productivity and capital intensity results for these categories are presented in table 3 below. This disaggregation makes clearer the main sources of the high productivity noted in section 3.1 above. 17 The table also includes a disaggregation across the categories for unit labor costs and labor costs per worker, which are discussed in section 3.4. TABLE 3: BREAKDOWN OF PARTIAL PRODUCTIVITY DATA BY FIRM TYPE Number of Labor Unit Labor Labor Costs Capital Workers Productivity Costs per Worker Intensity Small (5-19 workers) 10 $4,774 18% $1,071 $2,082 Medium (20-99 workers) 41 $2,905 25% $1,257 $1,487 Large (100 workers and 200 $8,855 16% $1,115 $9,700 up) Blantyre 20 $8,855 15% $1,190 $2,082 Lilongwe 19 $4,417 17% $1,071 $2,844 Non-exporter 18 $4,417 16% $1,071 $2,082 Exporter 161 $10,632 18% $1,741 $9,700 Domestic 15 $3,844 19% $1,062 $2,082 Foreign 150 $10,297 16% $1,611 $7,808 Credit constrained 18 $4,601 17% $1,171 $2,082 Not credit constrained 100 $7,928 12% $1,071 $4,183 Note: All measures are described in Annex 1. Groups with less than 10 observations are excluded from the table. Source: Authors’ calculation based on data from WBES 16 Further breakdowns with respect to manufacturing sub-categories are not possible at this point due to the small number of respondent firms in the respective sub-categories covered by the Enterprise Survey. 17 It will be noted that it is not possible to disaggregate the data further given the small number of firms in the different sub- sectors in manufacturing. 27 Partial Productivity with respect to Firm Sizes Larger firms in Malawi are on the whole more productive than small firms. Overall, labor productivity and capital intensity are significantly high among large firms, followed by small firms and medium sized firms (see Table 3 and Figure 8). The fact that large firms use much more capital intensive relative to other factors of production than smaller and medium firms is perhaps unsurprising. Interesting however, is the indication that medium sized firms are less capital intensive than small firms. This may mean they use more labor productive relative to capital productive, but it is noted that their labor costs per worker are also higher. The trend is also noted for labor productivity, which is lower for medium sized firms than small firms, and unit labor costs, which are higher for medium sized firms than they are for small firms and large firms, respectively. Higher labor costs are normally indicative of workers with higher skill levels being hired. Taken together, therefore, the above scenario of low labor productivity among medium sized firms may point at relevant skills not being available to them, at the same time that they use less capital relative to labor. These challenges in factors of production certainly reflect the “missing middle” that was highlighted in section 2.4, and require further interrogation to ensure the growth in medium sized firms. Of concern is the illustration in Figure 8, that labor productivity of medium-sized firms is less than that of both large and small firms. This is consistent with the “missing middle’ of firms highlighted in section 2.4. It may point to demand-side constraints on labor and needs to be interrogated further as it has implications on job creation. With regard to labor costs (see section 3.4), evidence from World Bank investment climate assessments generally shows that in most countries there is a gap in productivity between large and small firms, and that large firms also pay their workers more than small firms. For example, in Zambia, labor costs per worker are close to three times greater for large firms than they are for small firms. This is not the case in Malawi. Both large and small firms report labor costs close to US$1,000 per worker. This suggests that large firms are not passing the benefits of higher productivity to workers or are not hiring more skilled or better educated workers than small firms are. FIGURE 8: LARGE FIRMS ARE MORE LABOR PRODUCTIVE AND CAPITAL INTENSIVE $10,000 $8,000 $6,000 $4,000 $2,000 $0 Source: Authors’ calculation based on data from WBES Note: Partial productivity measures are measured in 2010 US$. See Annex for details. All data points are for the median firm on each measure of performance. 28 With respect to Location (Lilongwe and Blantyre) There are some notable differences in productivity across Malawi, with labor productivity much higher in Blantyre compared to Lilongwe (see Figure 9). The median firm in Blantyre produces about US$8,855 of output per worker. In comparison, the median firm in Lilongwe produces only US$5,112 per worker. On the other hand, firms in Lilongwe are more capital intensive than those in Blantyre. Whereas the median firm in Blantyre has US$ 2,082 of capital per worker, the median firm in Lilongwe has US$ 2,844 per worker. Unit labor costs are relatively similar at 15% and 17%, respectively (see table 3 above). FIGURE 9: BLANTYRE FIRMS ARE MORE PRODUCTIVE $8,855 $9,000 $8,000 $7,000 $6,000 $5,112 $5,000 $4,417 $4,000 $2,844 $2,844 $3,000 $2,082 $2,000 $1,000 $0 Blantyre All firms Lilongwe Labor productivity Capital intensity Labor Costs Source: Authors’ calculation based on data from WBES Note: Partial productivity measures are measured in 2010 US$. See Annex for details. All data points are for the median firm on each measure of performance. With respect to whether the firms engage in exports or produce for domestic market Firms that are export oriented are by and large more productive than those that are domestic oriented. A possible reason for higher productivity among firms that are export oriented is that these firms concentrate in capital intensive high skill sectors or that they have larger markets to work with and hence larger revenues for the same amount of labor. The median firm that is export oriented produces about $10,632 per worker, compared to US$4,417 per worker for a median firm that is not export oriented. While labor unit costs for the two categories are not too far apart - at US$1,000 and US$1,700 - the main difference between them is capital intensity. The export oriented firms use much more capital relative to other factors of production such as labor as compared to domestically oriented firms. (see Table 3). With respect to Credit Constraints Firms that have less credit constraints are more productive than those that are credit constrained. The reason seems to be that firms in this category have resources that make them access efficient machinery and are able to attain higher levels of operational efficiency, including getting more from the labor they employ. The median firm is more capital intensive, producing double output per available capital compared 29 to those that are credit constrained. Their labor productivity is much higher at US$7,928 per worker, compared to US$4,601 per worker for those firms that are credit constrained. (see Table 3). With respect to foreign or domestic ownership Foreign owned firms in Malawi tend to be more productive as they use labor and available capital more efficiently. In this analysis, the median foreign owned firm produced US$10,297 per worker compared to US$3,844 for the locally owned median firm. Similarly, a median foreign owned firm uses relatively more at US$ 7,808 per worker compared to US$2,082 for a locally owned median firm. These foreign owned firms are predominantly large scale, export oriented, in large cities such as Blantyre, and have better access to financing for their operations. 3.3 TOTAL FACTOR PRODUCTIVITY (TFP): CROSS COUNTRY ANALYSIS AND COMPARISON WITH RESPECT TO LOCATION IN MALAWI Partial productivity analysis can be misleading if considered in isolation. A firm could have high labor productivity because it is efficient, but it could also have high labor productivity simply because it uses a capital intensive production process. When multiple inputs are considered, such as labor and capital, the unaccounted for level of output compared to the level of inputs is commonly referred to as the Total Factor Productivity (TFP) or Multi Factor Productivity (MFP). TFP is calculated using regression analysis (For details about calculation and estimation of TFP, see Annex 3) . In this analysis, results (in comparison to other countries) are presented as a percentage in relation to Malawi. When the median firm’s TFP is higher in a country than in Malawi, TFP will be greater than 100 percent. When it is lower, TFP will be less than 100 percent. In theory, firms with higher TFP are more efficient because they produce higher output with fewer inputs. TFP measures how productive firms are based on technological progress, market structure and institutional effects, or other residual factors, after taking into account the firms’ use of both capital and labor.18 Typically, TFP would be higher when workers are better educated, more highly skilled, when managers are more efficient and when firms use advanced technology. Cross Country Analysis Analysis suggests that TFP is lower in Malawi than in most of the comparator countries . Although TFP is slightly higher for the median firm in Malawi than in Mozambique or Tanzania (57 and 77 percent of TFP in Malawi – see Figure 8), it is lower than in the other regional comparator countries. Although labor productivity in Malawi is higher than in Zambia, TFP is lower. This suggests more capital intensity in Malawi for example, than in firms in Zambia and Tanzania. Malawi’s reliance on tobacco processing might play a role, as that dominating sector is highly capital intensive. This might also be a reflection of Malawi’s low per capita income19. As TFP tends to be higher in countries where per capita income is greater, it is not surprising that TFP is relatively low in Malawi. Malawi might only compare favorably with other countries at similar levels of development. 18 See van Ark, B. (2014). Total Factor Productivity – Lessons from the Past and Directions for the Future. [Online. Available at: https://www.nbb.be/doc/ts/publications/wp/wp271en.pdf 19 Malawi’s per capita GNI was only $815 in purchasing power parity (PPP) US$ in 2014 (World Bank, 2015). This is lower than in all but three of the 188 countries where the World Bank has computed per capita GNI. 30 TFP with respect to Location A comparison is now drawn on TFP in different locations in Malawi using the same basic approach as was done to compare Malawi with other countries. Because the sample of manufacturing firms is relatively modest in Malawi—and because many manufacturing firms did not answer enough questions to calculate TFP — we calculate median levels of TFP not according to firm size but only with respect to three regions: Blantyre, Lilongwe and the rest of the country.20 MEDIAN TE RELATIVE TO 100% BLANTYRE 69% 59% Source: Authors’ calculation based on data from World Bank Enterprise Surveys (WBES) Note: The country estimates are from the LAD regressions. According to the TFP results above, firms in Blantyre are more productive than firms in other regions of the country. The median firm in Blantyre is about 41 percent more productive than the median firm in Lilongwe and about 31 percent more productive than the median firm in other regions. As the commercial hub of the country, Blantyre is home to a significant number of manufacturing firms. The higher output per worker could be reflective of competitive factor markets in which Blantyre participates, as well as its proximity to Malawi’s main trade routes through Mozambique and South Africa and hence greater access to raw materials and output markets in Malawi and outside. 3.4 LABOR COSTS AND PRODUCTIVITY: COUNTRY COMPARISON AND WITH RESPECT TO LOCATION AND FIRM SIZE Although TFP provides some useful incites about the competitiveness of firms relative to firms in other developing economies, it can be misleading when considered in isolation. One problem with TFP analysis is that it does not distinguish between highly skilled and unskilled workers. Productivity and wages will mostly be higher when firms employ highly skilled workers. However, firms can also remain competitive when productivity is low if wages are comparatively lower. For this reason, it would be useful to look at the cost and growth of labor as well as TFP when assessing competitiveness. The next sections therefore also consider labor costs, unit labor costs, and employment growth to assess competitiveness of enterprises in Malawi. 20 There were about 176 firms that were given the manufacturing survey. Many of these firms, however, were outside of manufacturing according to the ISIC codes that they gave during the interview. This can often happen when service or retail firms have some manufacturing (e.g. bakers). In total, of the 176 firms that were given the manufacturing survey, over half reported being in areas outside of manufacturing. In comparison, less than 2 percent of firms given the manufacturing survey in Zambia reported being in sectors outside of manufacturing. As a result, only 62 manufacturing firms answered enough questions to calculate TFP and reported doing business in areas that are considered manufacturing. 31 Steelworks production in a local firm Labor costs in Malawi are relatively high. Per worker labor costs are measured by dividing total expenditures on wages, salaries, and other benefits by the number of workers . The measure reflects payments to all workers including professionals and managers—not just production workers. For Malawi, they are estimated at US$1,071, which like labor productivity is higher than in neighboring Mozambique, Tanzania and Zambia, but less than Kenya, Zimbabwe and South Africa. As noted in table 3 and in the foregoing discussion, labor costs are highest among medium firms, followed by large and small firms. This suggests that large firms are not passing the benefits of higher productivity to workers and probably are not hiring more skilled or better educated workers than small firms are. This situation is not encouraging for the growth of small enterprises that are in majority in Malawi. $12,000 $9,673 60% 49% 45% $10,000 $8,000 40% 30% 27% $6,000 $3,900 21% 17% $4,000 20% 12% $871 $750 $913 $1,071$1,561 $2,000 0% Mozambique South Africa Tanzania Zimbabwe Zambia Malawi Kenya Mozambique Tanzania Zambia Malawi Kenya Zimbabwe South Africa Notes: 20% 16% 1. Partial productivity measures are measured in 15% 13% 2010 US$. See Annex for details. All data points 10% 5% are for the median firm on each measure of 5% 3% 3% performance 0% 2. Unit labor costs are the ratio of labor costs to labor productivity. See Annex for details Tanzania South Africa Mozambique Zambia Malawi Kenya Zimbabwe 3. Means not medians. See Annex for details on calculations. Source: Authors’ calculation based on data from World Bank Enterprise Surveys (WBES) Labor Costs and Productivity Unit labor costs allow for comparison of labor productivity and labor costs and how they might affect competitiveness. Unit worker labor costs are measured by dividing labor costs by unit of value added . Firms with high labor costs can still be competitive if they have relatively higher rates of labor productivity. On the other hand, unit labor costs are higher when high labor costs are not fully reflected in high productivity. When this is the case, firms will find it difficult to compete on international markets, all else equal. 33 Labor Costs Across the country and cities Unit costs in Malawi are relatively low. They are lower than most of the comparable countries, except Kenya (see figure 10). In countries where labor costs are lower than in Malawi, such as Tanzania and Mozambique, labor productivity is also lower. This suggests that Malawian firms might be able to remain competitive even though labor and total factor productivity is lower than in the best performing comparator economies. Although low salaries are not good for workers, they do allow firms to remain competitive when productivity is low, and can offset higher costs in other areas that are hard to reduce (such as the naturally higher transport costs associated with manufacturing in a landlocked economy). 3.5 EMPLOYMENT GROWTH Employment growth has been relatively modest in recent years in Malawi. Employment growth in the country averaged about 3 percent in the three years prior to the survey. This was lower than in any of the comparator countries except Zimbabwe, where employment growth was highly negative in the three years before that country’s 2011 survey. The status of job creation in and employment in respective sectors, challenges and suggested pathways are discussed in more detail in the 2017 Country Economic Memorandum for Malawi. In terms of demand, growth in manufacturing can create substantial scope for increased job creation. As different sectors grow and particularly as Malawi moves into production of higher value added goods and services, skills development that respond to current and emerging sectors is critical. 3.6 TOTAL FACTOR PRODUCTIVITY AND INVESTMENT CLIMATE An interesting question is how much TFP would improve if Malawi’s investment climate was more favorable. In this analysis, as indicated earlier, standard regression analysis is utilized to gain insights into this question. See Annex 3 for details on the methodology. Malawi’s investment climate is somewhat favorable, particularly after taking into account Malawi’s low per capita income - many investment climate variables tend to be consistently lower in poorer countries. The regression results (see table 4) illustrates that in the median country, firms spend less on bribes, are more likely to have bank credit and to have training programs, are more likely to license foreign technologies and are more likely to be foreign-owned than domestic firms in the median country. Comparisons are drawn with firms in Malawi at those in the 20th percentile (better than the median). Malawi does not compare favorably on a number of the measures, such as time spent dealing with regulation, losses due to crime and power outages, firms with bank credit and training programs and those likely to be exporters. This reaffirms some of the challenges that will be identified in chapters 4 to 6, which were based on the perceptions of firms. The gains that would accrue if Malawi improved towards levels of the median country and those in the 20th percentile, indicate the extent to which they affect TFP and would result in its increase if addressed. Firms’ TFP would increase only modestly if Malawi’s investment climate was as favorable as in the median country21 in all areas where Malawi’s performance does not already exceed the median country’s performance. The estimates suggest that if Malawi improved along all dimensions, average TFP would be about 10.4 percent higher than it is currently. The biggest gain would come from reducing crime—if losses due to crime were the same as in the median country, where TFP would be about 5.1 percent higher. Most of the other increases would be relatively modest. 21 Given that Malawi’s investment climate compares favorably with the median country’s investment climate, it is not surprising that the median manufacturing 34 TABLE 4: TFP AND IMPROVEMENT IN THE INVESTMENT CLIMATE Malawi Country Median TFP TFP at 20th Country Improvement Improvement if percentile if Malawi Malawi improved to improved to 20th median percentile Time dealing with regulation (%) 11.31 5.09 9.41 1.3% 0.4% Bribes (% of sales) 0.01 0.00 0.01 0.0% --- Losses to crime (% of sales) 1.85 0.13 0.35 5.8% 5.1% Losses due to power outages (% of sales) 0.78 0.01 0.11 1.7% 1.5% % of firms with bank credit 49% 78% 60% 3.3% 1.2% % of firms with own website 48% 62% 34% 4.5% --- % of firms that license foreign 26% 21% 14% --- --- technology % of firms with training programs 23% 49% 32% 3.3% 1.1% Firm is an exporter (dummy) 18% 45% 24% 5.2% 1.0% Firm is foreign-owned (dummy) 29% 17% 8% --- --- Total 25.1% 10.4% Source: Authors’ calculations based on World Bank Enterprise Survey data. Note: For variables that are entitled “% of firms” or “dummy variable”, the averages are weighted averages for manufacturing firms in the country. For the other (continuous) variables, the numbers are the weighted average response for manufacturing firms in the country. The median country (country at 20th percentile) is an unweighted median (country at 20th percentile) of country averages for that variable. Outliers more than 3 standard deviations are dropped for the averages for continuous variables with no upper bound (bribes, losses due to crime and losses due to power outages). See text for additional notes. If Malawi’s investment climate was equal to the best performing countries ranked in the 20 th percentile on each dimension, TFP would increase significantly by about 25.1 percent. In addition to the large gains from reducing crime, there would be additional gains from firms engaging more in exports (5.2 percent), increasing website use (4.5 percent), increasing access to bank credit (3.3 percent), increasing training programs (3.3 percent), decreasing power outages (1.7 percent) and time spent dealing with regulation (1.3 percent). 3.7 POLICY IMPLICATIONS OF PRODUCTIVITY MEASURES The foregoing discussion on productivity measures in Malawi highlights a number of implications for policy design and implementation, particularly with regard to developing the manufacturing sector: (a) Malawi’s higher labor productivity figures and lower unit labor costs as compared to its neighboring countries makes it competitive in attracting and channeling foreign and domestic investment into labor-intensive manufacturing industries. This is important for positioning Malawi to increase investment inflows vis-à-vis its neighboring countries, as well as address its low employment growth as highlighted in 8.3 above. While doing this, labor productivity needs to be further enhanced, particularly for medium-sized firms, as does their use of required capital. (b) The low capital productivity rates for Malawi as compared to all other countries in the region need to be addressed. The challenges in production due to low supply of electricity could be one way of ensuring greater utilization of capital. Further, increased earnings of income by individuals can create domestic demand for greater utilization of capital. Flowing from (1) above, an increase in labor-intensive industries can create jobs that can result in such increased incomes, thereby positively reinforcing the capital 35 utilization for domestic markets. Market access to regional and global markets can also enhance capital utilization as more goods are produced for those markets, of course along with the addressing of other supply-side constraints to production in Malawi. (c) After considering labor and capital, the TFP measure is one that requires attention if long term growth is to be sustained. As noted, TFP is only slightly higher than in Mozambique and Tanzania, but mostly lower than other regional comparator countries. As elucidated in Annex 4, there are a number of factors pertaining to the environment for investment that can be addressed to enhance TFP. These include: ▪ Intangible investments such as education and skills, research and development (R&D), patents, licenses, organizational change and product marketing; ▪ Skills, motivation and competencies – in part affected by intangible investments above; ▪ Innovation and technological change – in part also affected by the intangible investments above; and ▪ Markets, institutions and regulations. It is important that Malawi’s innovation ecosystem be effectively developed, coupled with frameworks that will encourage high levels of diffusion of technology. In this regard, R&D, education and skills development need to be emphasized, particularly as they have constrained productivity, growth and competitiveness thus far and are critical to enhancing it going forward. At the operational level, challenges have been noted in sectors such as agri-business, where progress has been constrained by lack of entrepreneurial skills, business organization, financial literacy, and other cognitive skills resulting from high illiteracy levels. Outside agriculture, skills development also has to respond to emerging sectors of high value added goods as alluded to in section 3.3 above. Progress in these areas needs to take into account the differences in productivity identified with respect to location as well as difficulties encountered according to gender of managers of firms highlighted in chapters 4 to 6. Specifically, drawing from the regression results presented in Table 4 above, firms are more productive when they spend less time dealing with regulation, have smaller losses due to power outages and crime, pay less bribes, have better access to credit and training programs and engage more in exports. These issues need to be addressed to increase productivity levels. On the whole, it will be noted that the favorable levels of productivity mostly reflected the position of large firms, which as mentioned earlier, are more able to navigate risks brought about by a difficult business environment. Efforts to raise productivity levels should, therefore, focus on facilitating the growth of small and medium sized firms. The next chapters explore the challenges to firms as reported in the enterprise surveys in more detail, highlight some efforts currently underway in different areas, and make policy recommendations for further improvement. 36 A bank pafupi customer CHAPTER 4: ENTERPRISE CONSTRAINTS: WHAT DO FIRMS IN MALAWI VIEW AS THE MAJOR CHALLENGES TO ENTERPRISE DEVELOPMENT AND GROWTH? The most frequently cited obstacle to business enterprises in Malawi is financing. This was identified as a major obstacle in the 2014 enterprise survey followed by electricity, corruption and tax rates. It was also cited as biggest obstacle in 2008 enterprise survey. FIGURE 13: MAJOR OBSTACLES IDENTIFIED IN PERCENTAGE SHARES Source: Authors’ calculation based on data from WBES 4.1 TOP THREE OBSTACLES BY FIRM SIZE, BUSINESS SECTOR, LOCATION AND GENDER PERSPECTIVE Enterprises in Malawi experience serious challenges to access financing. 29.9 percent of surveyed enterprises identified access to finance as the top most obstacle in the 2014 enterprise survey (was 45.7 per cent in 2008). A large group of respondents from the manufacturing sector (35.5 percent), identified this as the top most obstacle compared to 25.9 percent and 27.1 percent from retail and “other” services sectors, respectively. In terms of size of firms, a bigger proportion (37.3 per cent) of small scale enterprises perceived this as top most obstacle compared to medium scale (22.1 per cent) and large scale (17.1 per cent) enterprises. A bigger proportion (38.6 percent) were from firms that had female top management, compared to 28.7 percent from those with male top management. The majority of the firms were those that were producing goods primarily for the domestic markets. 38 TABLE 5: THREE BIGGEST OBSTACLES TO ESTABLISHMENTS Access to Electricity Corruption Finance All 29.9 14.2 10.7 Business Sector Manufacturing 35.5 17.0 6.1 Services 27.1 12.7 13.1 Retail 25.9 9.3 18.9 Other Services 27.7 14.7 9.9 Firm Size Small 37.3 9.2 12.0 Medium 22.1 13.8 7.2 Large 17.1 31.6 12.1 Location Blantyre 35.6 10.3 9.2 Lilongwe 21.7 20.3 14.4 Mangochi 13.9 18.4 0 Mzimba 17.6 7.5 4.0 Zomba 38.0 17.7 0 Status Exporters 19.8 19.7 18.1 Non exporters 28.4 14.3 11.5 Gender of Top Manager Top Manager, Female 38.6 11.4 0.2 Top Manager, Male 28.7 14.8 12.6 Ownership Domestic 37.5 8.3 11.4 More foreign 11.5 37.5 10.5 Source: Authors’ calculation based on data from WBES Access to Finance was a major concern in all three regions of the country. Between 17.6 and 37.8 percent of firms in the districts across the country indicated that this is a big obstacle for their firms. Most of those who held this view however were from cities of Zomba and Blantyre, and with a slightly lower percentage from Lilongwe. Electricity was the second biggest obstacle, and a major challenge especially to large enterprises. About 14.2 percent of enterprises in the 2014 survey indicated that electricity is the biggest obstacle that they faced. Large-scale firms (about 31.6 percent) reported this as their biggest obstacle, compared to 13.8 percent by medium scale enterprises and 9.2 percent by small scale enterprises. This points to some correlation between the size of firms and the electricity dependence in the firms’ operations, and hence the larger firms felt the impact more. Most of the respondents who cited electricity as the biggest obstacle are based in the cities of Lilongwe, Blantyre and Zomba, other than smaller urban towns such as Mangochi and Mzimba. Although firms in all regions were concerned about electricity, quite a significant proportion (20.3 percent) were from Lilongwe. Corruption was cited as the third biggest obstacle, and a major challenge mostly faced by the retail firms. Overall, 10.7 percent of enterprises indicated that this is the biggest obstacle they faced. About 12.0 percent and 12.1 percent of both the small and large-scale firms respectively, reported this as the biggest obstacle to their businesses, while only 7.2 percent of medium-sized firms cited corruption as the biggest obstacle. Corruption was reported as the biggest obstacle particularly by retail firms (18.9 percent), followed by services firms (13.1 percent), and manufacturing firms (6.1 percent). Lilongwe firms cited corruption as the biggest obstacle in comparison with Blantyre and other places. The problem seems to be more prevalent in cities other than the rural urban areas. 39 4.2 COMPARISON WITH 2008 SURVEY Access to finance was the most commonly cited concern in both the 2008 and the 2014 enterprise surveys. About 45.7 percent of firms reported that access to finance is the biggest constraint in the 2008 survey, but this reduced to 29.9 percent in the 2014. In both surveys, there is a large gap between access to finance and the next obstacle. This gap, however, narrowed, in the 2014 survey findings, reflecting a decrease in perception of access to finance as a major obstacle and an increased concern for power outages. FIGURE 14: TOP TEN BIGGEST OBSTACLES TO ENTERPRISES IN MALAWI, 2008 AND 2014 Access to finance 29.9 45.7 Electricity 14.2 Corruption 10.7 Tax rates 10.1 Practices of the informal sector 7.6 Access to land 7.1 Political instability 4.2 Business licensing and permits 3.6 Crime, theft and disorder 3.2 Transportation 2.7 0.0 10.0 20.0 30.0 40.0 50.0 Malawi 2014 Malawi 2008 Source: Authors’ calculation based on data from WBES In the 2014 enterprise survey, the second on the list as top most obstacle was electricity outages, while in 2008, transportation was second. 14.2 percent of firms view power outages as the biggest challenge in 2014. This represents an increase from the 8.6 percent in 2008. The proportion of firms that cited transportation as their biggest obstacle reduced from 11.4 percent in 2008 to 2.7 percent in 2014. This is the most notable difference in the two survey results and may be due to progressive improvements in the quality of both national and regional transport infrastructure over the years. Corruption has been on the rise in the country and was significantly reported to be rampant at the time of the Enterprise Survey. This became third on the list of obstacles to enterprises in 2014, replacing practices of the informal sector, which was third on the list of obstacles in 2008. 10.7 percent of respondents chose corruption as the top most obstacle, up from a much less prominent 2.7 percent in the 2008 survey. 40 4.3 CROSS COUNTRY COMPARISON ON TOP MOST OBSTACLES TABLE 6: ACCESS TO FINANCE, ELECTRICITY AND CORRUPTION FIGURE 15: TOP THREE OBSTACLES - COUNTRY COMPARISON Access Electricity Corruption to Finance Tanzania (percentages) Malawi Malawi 29.9 14.2 10.7 Zambia Sub-Sahara 21.9 12.5 8.0 Mozambique Zimbabwe Zambia 27.5 13.1 2.5 Kenya Zimbabwe 20.6 3.3 8.3 South Africa Kenya 9.6 9.6 12.3 Sub−Saharan Africa Tanzania 37.9 24.9 2.5 0 10 20 30 40 South Africa 7.5 14.7 7.1 Mozambique 23.2 9.1 4.1 Source: Authors’ calculation based on data from WBES Evidence from enterprise surveys suggests that access to finance is a problem in most Sub-Saharan African countries. The percentages of firms that reported access to finance as biggest obstacle in Malawi is higher than the regional average. Access to finance is also a top obstacle in Tanzania, Zambia, Mozambique and Zimbabwe. This is however not the case in Kenya where corruption is the biggest obstacle and South Africa where electricity is the biggest obstacle. Access to finance and electricity are the first and second biggest obstacles for firms in Sub-Saharan Africa. 4.4 COMPARISON WITH THE WORLD BANK DOING BUSINESS SURVEY RESULTS Drawing comparisons between the DB 2015 report and the ES 2014, which were conducted around the same time, shows similarities in the key challenges facing businesses in Malawi for example in access to finance, electricity and getting services from Government institutions. See the table below for details. A summary of the ES and DB methodologies is in Annex 6. 41 Comparison of Findings in DB and ES22 Indicator Doing Business Enterprise Survey Access to Ranking of the ‘Getting Credit’ Access to Finance was the biggest Finance indicator worsened in the 2015 DB Obstacle to enterprises in Malawi report to 151st from 147th in 2014. Electricity The ‘Getting Electricity’ indicator Electricity was perceived by the ranked 181st in 2015, which was the largescale enterprises as the biggest worst ranking across all indicators. obstacle to business development in The 2015 DB report indicated that the Enterprise survey. There is a it took 222 days to get connected to difference in the findings of the the electricity grid, compared to specific number of days to get 109 days in Tanzania, 34 days in connected to the electricity grid, but Rwanda, and 107 days in both findings indicate that the Mozambique. severity of the electricity challenge in Malawi. Paying Malawi’s ranking deteriorated from Tax rates were ranked 4th biggest Taxes 92nd in the 2014 DB report to 103rd obstacle by enterprises in 2014, and in 2015. 10 percent indicated that this is a major problem (compared to 7.0 percent in 2008. Similarly, larger percentage of firms mentioned tax administration as a major obstacle, compared to 9.0 per cent in 2008.) Trading The DB indicated that the time that In the ES, firms reported that they Across it took to export and import was were expected to give gifts in order Borders significantly higher than to get necessary documentation comparator countries. Similarly, such as import licenses (see table the amount of documentation was 14) more than in comparator countries. Dealing with The DB indicated that it took 60 In the ES, businesses reported that Construction days to get a construction permit they were expected to give gifts to permits get a construction permit (see table 14) Source: Source Enterprise Survey (World Bank, 2017c) 4.5 COMPARISON WITH THE MCCCI BUSINESS CLIMATE SURVEY In 2015, the Business Climate Survey by the Malawi Confederation of Chambers of Commerce and Industry (MCCCI), asked business owners/leaders the extent to which they perceived several factors as challenges to doing their businesses in Malawi. Participants were asked to rate challenges on a scale of 1 to 10 (1 to 5 being minor, 6 to 7 being moderate and those above 7 considered as major obstacles). The findings are similar to those from the 2014 enterprise survey, and also consistent with the DB findings. Cost of finance was cited as one of the top three obstacles. This comes from high interest rates in the country as a result of the persistently high inflation and weak productive structure. The other big obstacles were telecommunication, electricity, economic uncertainty and policy inconsistency (see figure 16). 22 Note: Most of the included indicators have improved in the 2016 and 2017 DB Reports on account of some reforms implemented in the respective areas. For example, the ‘Getting credit’ indicator has improved to 101 in 2017, following implementation of the Collateral Registry in 2016. 42 FIGURE 16: TOP OBSTACLES ACCORDING TO SURVEY BY MCCCI Cost of Finance remains biggest obstacle, followed by electricity Cost of Finance 9.6 Telecommunication 9.1 Electricity 8.7 Economic uncertainity & Regulatory Policies 8 Water 7.6 Crime 7.6 Parliament effectveness 7.4 Customs regulations, procedures,… 7.3 Taxes 7.1 8 10 12 Source: Malawi Confederation of Chambers of Commerce and Industry (2016) 43 CHAPTER 5: LOOKING AT TOP CONSTRAINTS IN DETAIL In Chapter 2, it was noted that the 2014 enterprise survey took place against a backdrop of pronounced economic and financial sector challenges. The banking sector was dominated by a few large commercial banks commanding a majority loan portfolio. There were few lending instruments, mostly on commercial terms without long term development finance instruments. Macroeconomic conditions were significantly adverse. Inflation was very high, at 25.9 percent in January 2014. Monetary conditions were tight with policy rate above the 25 per cent. This translated into significantly high commercial bank interest rates of above 40 per cent, which were largely unaffordable to the common enterprise firms. This, together with a largely undeveloped credit system, made financial institutions reluctant to lend out funds, especially to small scale enterprises. The financial sector was not well developed (without interoperability and limited transactional innovations such as e-money or e-banking), making it difficult for those in rural areas to access financial services. Analysis of the 2014 enterprise survey results confirms these challenges and provides insights on the profile across sectors and demographics - showing for example that access to finance is a bigger challenge for service industries, small scale enterprises, rural based, female headed, non-exporting, and domestic oriented firms. Few firms have access to loans, but this challenge is probably less problematic for those firms that are retailers and engaging in exports. This is also the case for firms with foreign ownership. Malawi is amongst the countries that have the highest percentage of loans requiring collateral and also highest loan-to-collateral-value ratios. Only a small proportion of working capital in Malawi is financed by banks, and this is mostly to the largescale, the manufacturers, the export oriented and the foreign owned firms. As a result, the majority of the firms in the country depend heavily on funding their investments through own funds. Only a third use banks to finance investment. This chapter explores questions related to access to finance in more detail. Chapter 4 focused on the question requesting respondents to indicate an element from a list, that represents the biggest obstacle to their enterprises. The analysis in the next section of this chapter focuses on the degree23 to which an element of business environment is an obstacle, starting with the key challenges identified by the Enterprise Survey. 23 Whether to a “minor” or “major” degree 44 5.1 ACCESS TO FINANCE IS A MAJOR CONSTRAINT TABLE 7: FINANCE AS A MAJOR OBSTACLE (percentage) FIGURE 17: ACCESS TO FINANCE AS MAJOR CONSTRAINT - 2008 2014 COUNTRY COMPARISON Access to Finance as biggest obstacle 45.7 29.9 Percentage of firms identifying Access 51.0 34.9 to Finance as a major constraint Sector Zimbabwe 55.9 Manufacturing 48 33.7 Services 52.1 35.6 Retail 33.4 Mozambique 50.1 Other Services 36.7 Size Tanzania 43.9 Small 54.2 40.0 Medium 52.9 33.4 Malawi 34.9 Large 30.5 19.4 Location Zambia 27.4 Blantyre 32.1 Lilongwe 38.1 Kenya 17.2 Mangochi 52.9 Mzimba 28.7 South Africa 15.5 Zomba 56.4 Exporter (portion of sales are exports) 47.9 26.7 Sub−Saharan Africa 37.8 Domestic only firm (no exports) 51.5 37.0 Top Manager, Female 64.7 41.9 0.0 10.0 20.0 30.0 40.0 50.0 60.0 Top Manager, Male 48.5 33.6 Firm has no foreign ownership 57.6 43.1 Firm has at least 10% foreign ownership 30.3 16.3 Source: Authors’ calculation based on data from WBES Access to finance was cited as the biggest obstacle that firms face in their daily operations in Malawi and as a major constraint. In response to a question of what managers consider, out of a list of 15, as the biggest obstacle to their establishments, 29.9 percent indicated that access to finance is the biggest obstacle (from 45.7 per cent in 2008). Further, when asked to what degree access to finance is an obstacle on a scale of 1 to 424, 34.9 percent responded that this was a major constraint. Malawi is however, not the worst amongst comparable economies. This finding is better if compared with the average for Sub-Saharan countries, and it represents an improvement from 2008 when 51.0 percent identified access to finance as a major constraint. Larger proportions of firms in Zimbabwe (55.7 percent), Mozambique (50.1 percent) and Tanzania (43.9 percent) also identified this as a problem. Access to finance seems to be a bigger challenge for service industries. 35.6 percent of service enterprises identified access to finance as a major constraint compared to 33.7 percent for firms engaged in manufacturing. Similarly, 40.0 percent of small scale enterprises identified access to finance as a major obstacle compared to 33.4 percent and 19.4 percent for medium scale and large scale enterprises, respectively. A proportion of manufacturing and service firms that indicated access to finance as a challenge reduced to a third in 2014 compared to around half in 2008. 24 Thequestion was asked on a basis of scale 1 to 4 (1 = no obstacle, 2 = minor obstacle, 3 = moderate obstacle, 4 = very severe obstacle) 45 The rural based, the female headed, the non-exporting, and the domestic oriented firms seem to face more challenges with access to finance than their respective counterparts. A higher proportion of firms in Mangochi (52.9 percent) identified this as a major constraint than from the more urban cities of Lilongwe (38.1 per cent) and Blantyre (32.1 percent). A significant number of firms in Zomba (56.4 percent) also cited this challenge. Of the total that cited access to finance as a major constraint, 41.9 percent were firms with a female top manager against 33.6 percent from firms with a male top manager. 43.1 percent were firms with no foreign ownership while 16.3 percent were those which had at least 10 percent foreign ownership. 37 percent were firms selling to the domestic markets only, while 26.7 percent were those that exported some portion of their sales. Access to Loans Access to credit/loans remains low in Malawi. Only 26.7 percent of firms interviewed in the 2014 enterprise survey indicated that they have a loan or line of credit with a bank. This is low compared to 2008 (40.1 percent), but higher than the average for Sub-Sahara (22.6 percent) and neighboring countries: Tanzania (16.6 percent), Mozambique (14.2 percent percent) and Zambia (8.8 percent).25 TABLE 8: FIRMS WITH CREDIT LINES (percentage) 2008 2014 FIGURE 18: FIRMS OWNING BANK CREDIT LINE /LOAN - COUNTRY Proportion with bank credit line 40.1 26.7 COMPARISON Small scale 27.7 21.3 Medium scale 59.9 30.0 Large scale 51.9 40.2 Business Sector Zimbabwe 7.8 Manufacturing 42.5 25.8 Retail n.a 28.0 Zambia 8.8 Services 39.3 27.2 Exporter (portion of sales are 51.5 53.3 Mozambique 14.2 exports) Domestic only firm (no exports 40.0 23.9 Tanzania 16.6 Ownership Top Manager, Female 37.9 38.7 Malawi 26.7 Top Manager, Male 40.6 24.9 South Africa 30.1 No foreign ownership 40.3 24.9 At least 10% foreign 39.5 25.3 Kenya 35.6 ownership Location Sub−Saharan Africa 22.6 Lilongwe 29.0 Blantyre 25.8 10 20 30 40 Mangochi 9.2 Mzimba 22.4 Zomba 28.3 Source: Authors’ calculation based on data from WBES The large, the retail, the export oriented, and the firms with foreign ownership are more likely to have a bank loan or line of credit. Access to loan facilities is skewed towards the large-scale enterprises (40.2 percent), compared to the medium (30.0 percent) and the small scale enterprises (21.3 percent). There are also more retail firms (28.0 percent) that reported to have such a facility than the manufacturing firms 25 Given the tough macro-economic environment some/most of the firms surveyed in 2008 may have closed (fallen out) with less new fims formed, contributing to the lower percentage(what is termed ‘Survivorship Bias‘).If more new firms were formed, then the new are less likely to access credit from financial institutions 46 (25.8 percent). Similarly, it is skewed towards the firms engaged in exports (53.3 percent) compared to those with domestic orientation (23.9 percent). The firms that are headed by female top managers still reported having more access to bank credit than their counterparts, as did those firms with foreign ownership. Those firms located in urban areas such as Lilongwe and Blantyre reported as having more access to a line of credit than in the rural urban centers such as Mangochi. Collateral Requirements Malawi is amongst the countries that have the highest percentage of loans requiring collateral and highest collateral to loan value ratio. In the 2014 enterprise survey, 92.6 percent of loans in Malawi required collateral (see Table 9). This compares unfavorably with results from the 2008 enterprise survey which indicated that 88.1 percent required collateral. The value of collateral needed for a loan is equally large at 293.6 percent of the value of a loan. This translates to over three hundred percent above 2008 level and is the highest in the region. TABLE 9: COLLATERAL REQUIREMENTS Proportion Value of FIGURE 19: LOANS REQUIRING COLLATERAL -COUNTRY COMPARISON of loans collateral requiring needed collateral for a 350.0 293.6 (92.6 %) loan 300.0 (293.6 250.0 213.2 223.4 236.6 240.2 %) 188.4 Firm Size 200.0 Small scale 94.9 316.8 150.0 Medium scale 97.6 345.8 92.0 103.6 92.6 Large scale 82.7 204.8 100.0 Business Sector 50.0 Manufacturing 99.1 229.6 Retail 89.7 346.3 0.0 Services 89.2 340.8 Exporter (portion of sales 97.4 160.7 are exports) Domestic only firm (no 96.2 308.2 exports Gender of Top Manager Top Manager, Female 100 360.7 Top Manager, Male 90.7 274.6 Value of collateral needed for a loan (% of the Ownership loan amount) No foreign ownership 96.3 311.5 Firm has at least 10% 97.1 247.0 Proportion of loans requiring collateral (%) foreign ownership Location Lilongwe 92.6 324.1 Linear (Proportion of loans requiring collateral Blantyre 92.3 286.4 (%)) Mzimba 100 240.1 Source: Authors’ calculation based on data from WBES The Value of collateral in Malawi and comparator countries is very high despite low bank non- performing loans to total gross loans (NPLs) . For example, the NPLs for Mozambique for the year that the Enterprise Survey was conducted was 2.6 and the value of collateral needed for a loan was 92 per cent of the loan amount (Figure 20). For Tanzania, the NPLs was only 5.0 but the value of collateral was as exorbitant as 240.2 per cent. For Malawi, the value of collateral was 293 % of loan value, but the NPL to 47 ratio was 15.6%. This evidence suggests very little, if at all any, strong relationship between value loan requirements and NPLs26. FIGURE 20: COMPARISON OF VALUE OF COLLATERAL AND NPLS – MALAWI AND COMPARATOR COUNTRIES IN YEAR OF ENTERPRISE SURVEY 293.0% 300.0% 236.6% 240.2% 250.0% 188.4% 200.0% 150.0% 92.0% 103.0% 100.0% 0.5 50.0% 6.9% 5.1% 15.6% 2.6% 1.4% 0.0% NPL Value of Loan Requirements Source: Authors’ calculation based on data from WBES and the WDI (World Bank, 2017b) Complaints about the high proportion of loans requiring collateral and high value of collateral requirements seem to cut across all firms irrespective of size, sector, location, and whether domestic or foreign oriented. One interesting observation is a gender issue, with more reported disadvantages for firms and enterprises whose managers are women. For example, the value of collateral needed for a loan from such firms were at 360.7 percent of loan value, against 274.6 percent of loan value for firms headed by males. The Personal Property Security Registry system implemented by the Government in 2016 with support from the World Bank provides a good opportunity for countering this situation, as business owners can now use their movable assets to secure loans27. Working Capital A sizable number of firms, especially large scale in size, exporting, with foreign ownership and those located in cities, utilize banks more as a source of financing for working capital. Almost a third (29.2 percent) of firms surveyed reported that they used banks to finance working capital. 42.9 percent of large scale enterprises utilize banks this way compared to 22.0 percent of small-scale enterprises. 30.3 percent of firms in the manufacturing sector also use banks as a source of financing for working capital compared to 22.6 percent for retail firms. Firms with foreign ownership utilizing this facility for working capital more than those with only domestic ownership. An interesting aspect is that a higher percentage of firms with 26 Available data shows that NPL in Malawi jumped from around 2.0 per cent in early 2012 to 13.6 per cent in late 2013. And the value of collateral jumped from 161.1 per cent of vale of loan to 293.6 per cent in 2014, according to RBM Stability Report and Enterprise survey) 27 http://www.worldbank.org/en/news/feature/2016/02/04/new-online-collateral-registry-system-facilitates-increased-access-to- business-loans-in-malawi 48 female top managers indicated utilizing this feature for working capital compared to those firms headed by male top managers. TABLE 10: WORKING CAPITAL FINANCED BY BANKS Working Percentage of firms FIGURE 21: PERCENTAGE OF FIRMS USING BANKS TO capital using banks to FINANCE WORKING CAPITAL financed by finance working banks capital (29.2%) (10%) Small scale 7.6 22.0 Kenya 41.1 Medium scale 14.0 41.7 Large scale 14.4 42.9 Malawi 29.2 Manufacturing 11.1 30.3 Retail 7.0 22.6 South Africa 21.1 Services 9.3 28.6 Exporter (portion of 16.3 48.8 Tanzania 14.7 sales are exports) Domestic only firm 9.9 27.9 (no exports Zimbabwe 12.0 Top Manager, 13.0 36.9 Female Zambia 9.9 Top Manager, Male 9.5 27.9 Firm has no foreign 9.3 24.7 Mozambique 8.5 ownership Firm has at least 13.5 48.4 10% foreign Sub−Saharan Africa 23.7 ownership Lilongwe 13.3 36.4 0 10 20 30 40 50 Blantyre 8.3 26.1 Mangochi 2.3 16.2 Mzimba 19.7 45.2 Source: Authors’ calculation based on data from WBES Financing of working capital by banks is skewed towards the largescale, the manufacturers, the export oriented, the female headed, and the foreign owned firms. 14.4 percent of working capital was financed by banks for the large scale enterprises compared to 7.6 percent for the small scale enterprises. For the manufacturers, 11.1 percent of firms were financed by banks compared to 7.0 percent for the retailers. 16.3 percent of working capital was financed for the export oriented firms, compared to 9.9 percent for the domestic oriented firms. 13.0 percent of working capital for the firms with female top managers was financed by banks, against 9.5 percent for their male counterparts. 13.5 percent of working capital was financed for the firms with some foreign ownership component against 9.3 percent for those without. Investments The firms in Malawi depend heavily on funding their investments through own funds, with occasional recourse to other sources. 65.8 percent of total investment is financed internally, and the rest by banks, with little amounts coming from supplier credit, equity or stock sales (Figure 23. While in small proportions, there is some growth in investment financing through supplier credit, and equity or stock sales, compared to 2008. Between 2008 and 2014 the proportion of investments financed by supplier credit increased from 2.5 percent to 7.4 percent and proportion for investment financed by equity increased from 2.9 percent to 7.7 percent. The proportion of investments financed internally reduced from 75.4 percent to 65.8 percent. 49 A third of the firms in the country use banks to finance investment, and this proportion compares well in the region. The 2014 enterprise survey results indicate that 30.3 percent (compared to 20.6 percent in 2008) of firms in Malawi use bank financing for investment (Table 9). Whilst this is higher than the average for Sub-Sahara, it is better than the neighboring and comparable economies such as Zambia (12.2 percent), Tanzania (18.5 percent), Mozambique (10.5 percent), and Zimbabwe (16.7 percent). It is, however, lower than Kenya (43.2 percent), and South Africa (34.8 percent). TABLE 11: FIRMS USING BANKS TO FINANCE INVESTMENT Percent of FIGURE 22: FIRMS USING BANKS TO FINANCE INVESTMENT (%) firms using banks to finance investments Kenya 43.2 (30.3%) Size South Africa 34.8 Small scale 23.5 Malawi 30.3 Medium scale 34.5 Tanzania 18.5 Large scale 42.0 Sector Zimbabwe 16.7 Manufacturing 29.9 Zambia 12.2 Retail 28.8 Services 30.6 Mozambique 10.5 Exporter (portion of 52.0 Sub−Saharan Africa 21.3 sales are exports) Domestic only firm 26.1 10 20 30 40 50 (no exports Management and Ownership FIGURE 23: INVESTMENT FINANCING OPTIONS (PROPORTION, %) Top Manager, Female 31.2 Top Manager, Male 30.5 Firm has no foreign 26.7 65.8 80 ownership Firm has at least 10% 39.2 60 foreign ownership Location 40 Lilongwe 27.8 13.8 20 7.7 7.4 Blantyre 30.2 Mangochi 7.5 Mzimba 50.3 Internally By Banks By By Supplier Source: Authors’ calculation based on data from WBES The largescale, the export oriented and the foreign owned firms seem to have more access and use banks to finance investments than their local small scale domestic oriented counterpart. 42.0 percent of the large scale enterprises use banks to finance investment against 34.5 percent and 23.5 percent of the medium scale and the small scale enterprises, respectively. 52.0 percent of the firms that export portions of their sales use banks to finance investment, against 26.1 percent for the domestic oriented enterprises. 39.2 percent of the firms with at least 10 percent of foreign ownership use bank financing for investment against 26.7 percent for those with no foreign ownership. 50 5.2 ELECTRICITY OUTAGES It was noted in the contextual background that only about 9 percent of the population has access to electricity in Malawi. Demand for power has been growing, and as indicated the number of customers in 2015 was 312,857 while the generation capacity remained at 350MW (as in 2014), against the demand of 440 MW. There are a number of other challenges in the sector apart from the low installed capacity, against rising demand and the reliance on Shire River for generation 28. These include inadequate transmission and distribution networks, low tariff rates affecting economic viability of new projects, as well as uncertainties in the regulation and governance of the sector. Analysis from the 2014 enterprise survey shows the extent to which electricity outages have affected businesses in Malawi, with it emerging as their second biggest obstacle . In Chapter 3, it was noted that about 14.229 percent of top managers interviewed in the 2014 enterprise survey considered electricity outages as the biggest obstacle to their enterprises. In a typical month, firms reported experiencing 6.7 outages, each lasting an average of 3.5 hours. On a scale of 1 to 4 concerning severity of the problem, 24.8 percent identified this as a major obstacle. This chapter analyzes this challenge and compares it with previous survey results, and situations in neighboring countries. Electricity outages are a big challenge in Malawi, although this is better than in some of the comparator countries. According to the enterprises survey, Tanzania experiences the largest number of outages in a typical month (8.9 outages a day), followed by Malawi (6.7), Kenya (6.3) and Zambia (5.2). Outages in Tanzania, Zambia and Kenya were reported to have lasted longer than in Malawi. Kenya reported experiencing biggest losses as a percentage of annual sales due to the electricity outages (Figure 24). As a coping mechanism, firms turn to generators to make up for the lost energy. About 40.9 percent of firms in Malawi own or share a generator. This is less than the average for Sub-Sahara (50.9 percent), and also less than Zimbabwe (62.3 percent), Kenya (57.4 percent) and Tanzania (43.0 percent). But it is worse than Zambia, South Africa and Mozambique. With wide usage of power back-up systems, the average cost of power in Malawi and other countries in the region is high. While Malawi’s grid electricity is generally recognized to be reasonably low-cost, this is offset at the enterprise level by the cost of investing in and running back-up generating capacity. 28 Malawi is still not connected to the Southern Africa Power Pool (SAPP), therefore not yet able to trade in electricity through the SAPP. 29 Slightly higher than the Sub-Saharan Africa average (13.7 percent) 51 FIGURE 24: ELECTRICITY OUTAGES AND ASSOCIATED LOSSES ACROSS COUNTRIES 8.9 8.3 FIGURE 25: FIRMS OWNING OR SHARING 6.7 GENERATORS (%) 6.3 5.2 4.5 1.6 62.3 0.9 Zimbabwe Kenya 57.4 Sub−Saharan Africa South Africa Kenya Zambia Malawi Zimbabwe Tanzania Mozambique Tanzania 43.0 Malawi 40.9 Zambia 27.3 South Africa 18.4 Mozambique 12.6 Sub−Saharan Africa 0.0 20.0 40.0 60.0 80.0 Source: Authors’ calculation based on data from WBES Comparison with 2008 Survey There were more electricity outages in Malawi in 2014 compared to 2008. 24.8 percent of the firms in the 2014 enterprise survey identified electricity as a major constraint, compared to 37.6 percent in the 2008 survey, but firms reported 6.7 outages in a typical month in 2014, compared to only 0.8 outages in 2008 on average. A typical outage lasted longer (3.5 hours) in 2014 compared to 2.4 hours in 2008. This is a drastic deterioration that does not bode well for the operations of enterprises and their continued growth. The electricity outages translate into significant losses to industry revenue in Malawi. About 5.1 percent of annual sales were reported to have been lost due to electricity outages in 2014, down from 8.0 percent in 2008. This decline in losses can probably be explained by an increase in the usage of generators, as the number of firms owning generators almost doubled (from 25.3 percent in 2008 to 40.9 percent in 2014) (Table 10). 52 TABLE 12: FIRMS IDENTIFYING ELECTRICITY AS MAJOR CONSTRAINT 2008 2014 FIGURE 26: NUMBER OF DAYS TO OBTAIN ELECTRICITY % of firms identifying electricity as 37.6 24.8 CONNECTION a major constraint Number of outages in a typical 0.8 6.7 month Tanzania 52.6 Large 1.0 9.4 Medium 0.9 7.5 Malawi 50.4 Small 0.7 Na Duration of typical outages 2.4 3.5 43.0 (hours) Large 4.7 94 Zimbabwe 22.2 Medium 2.7 7.5 Small 1.8 5.7 Zambia 18.9 Losses due to outages (% of 8.0 5.1 annual sales) South Africa 15.8 Large 5.6 3.2 Medium 15.3 7.3 Mozambique Small 4.8 4.6 Firms owning generator (%) 25.3 40.9 Sub−Saharan Africa 36.8 Large 46.0 74.0 Medium 34.4 60.5 20 40 60 Small 4.1 23.1 Source: Authors’ calculation based on data from WBES There were more electricity outages in Malawi in 2014 compared to 2008. 24.8 percent of the firms in the 2014 enterprise survey identified electricity as a major constraint, compared to 37.6 percent in the 2008 survey, but firms reported 6.7 outages in a typical month in 2014, compared to only 0.8 outages in 2008 on average. A typical outage lasted longer (3.5 hours) in 2014 compared to 2.4 hours in 2008. This is a drastic deterioration that does not bode well for the operations of enterprises and their continued growth. The electricity outages translate into significant losses to industry revenue in Malawi. About 5.1 percent of annual sales were reported to have been lost due to electricity outages in 2014, down from 8.0 percent in 2008. This decline in losses can probably be explained by an increase in the usage of generators, as the number of firms owning generators almost doubled (from 25.3 percent in 2008 to 40.9 percent in 2014) (Table 10). Malawi firms face challenges to get connected to the electricity grid. Firms reported that it takes 50.4 days to obtain an electrical connection (upon application) in Malawi. While this is a slight improvement compared to 59.2 days in 2008, it is still very high and is worse than some of the countries in the region such as Kenya (43 days), Zimbabwe (22.2 days), Zambia (18.9 days), and Mozambique (12.7 days). It is only better than Tanzania at 52.6 days (see Figure 26). Perceptions with Respect to Size of Firms Large scale and medium scale enterprises were more concerned about electricity outages in 2014, compared to 2008. In the 2014 Enterprise Survey, large scale enterprises reported 9.4 outages in a typical month as compared to one outage in a typical month in 2008. Medium scale enterprises reported 7.5 outages in 2014 compared to 0.9 in 2008. 53 An increasing number of firms in Malawi turn to generators as a source of complementary energy . 74.0 percent of the large scale enterprises and 60.5 percent of medium scale enterprises reported owning or sharing a generator compared to only 46.0 and 34.4 percent in 2008. A lesser share of small-scale enterprises own generators, at 23.1 percent in 2014 (up from 4.1 percent in 2008). Medium size firms seem to be affected more from electricity outages than large scale and small scale firms. The medium sized firms reported losing 7.3 percent of annual sales due to electricity outages in 2014, compared to 4.6 percent of annual sales for small scale enterprises and 3.2 percent for large scale enterprises. TABLE 13: POWER OUTAGES IN A TYPICAL YEAR IN 2008 AND 2014 Number of electrical Number of electrical outages in a typical month outages in a typical month 2008 2014 Firm Size Small 0.7 5.7 Medium 0.9 7.5 Large 1.0 9.4 Firm Type Manufacturing 0.8 7.7 Service/retail 0.8 5.5 Type Exporter 0.8 8.0 Non exporter 0.8 6.2 Ownership Type Male Top Manager 0.7 7.1 Female Top Manager 1.1 4.8 Domestic 0.9 6.4 Foreign 0.6 7.8 Location Blantyre 5.9 Mangochi 10.6 Lilongwe 7.8 Mzimba 6.7 Zomba 8.2 Source: Authors’ calculation based on data from WBES Perceptions with Respect to Business Sector Manufacturing firms are affected more strongly by the electricity outages in Malawi. Firms engaging in manufacturing reported in the survey that they experienced 7.7 electricity outages in a typical month, while those engaged in retails and general service sector reported an average of 5.5 outages and 6.6 outages in a typical month, respectively. Prospects for increased manufacturing growth in the Malawi economy are evidently undermined by the outages, and partly by high usage of generators. As a result, manufacturing firms reported more losses due to this challenge than service and retail firms. Manufacturing firms reported losses of 6.1 percent of annual sales, while service firms reported 4.7 percent, and retail firms reported 4.3 percent losses of annual sales. 47.2 percent of manufacturing firms reported that they owned or shared a generator in 2014 54 and 27.8 percent of their electricity was from the use of generators. This is compared to 39.1 percent of service firms and 35.2 percent of retail firms that owned or shared a generator. Perceptions with Respect to Location Electricity outages seem to be a challenge in both the rural and urban centers. Electricity outages were reported at between 5.9 and 10.6 in a typical month across the surveyed districts. These outages were on average between 3.8 and 4.6 hours in the major cities of Blantyre, Zomba and Lilongwe. Losses due to electricity outages were slightly lower in the major cities. Blantyre firms reported losses of 3.6 percent of annual sales, while Lilongwe reported 6.2 percent. Meanwhile, firms in Zomba reported losses of 13.8 percent of annual sales, followed by Mangochi (12.9 percent) and Mzimba (11.8 percent). Firms in the capital city seem to be more constrained than their Blantyre counterparts. 51.6 percent of firms in Lilongwe reported owning or sharing a generator, and that 35 percent of their electricity came from generators. Perhaps due to the lower frequency and shorter durations of outages, only 31.1 percent of Blantyre firms reported to own or share a generator, and that only 18.9 percent of their electricity comes from generators. 5.3 CORRUPTION Corruption is seen as the third most significant obstacle to doing business in Malawi. The 2014 Enterprise Survey was conducted at the time of revelations of ‘Cashgate’ – a financial scandal that involved systematic syphoning of public resources through unperformed contracts. Increased corruption – bribery in particular – was validated as an obstacle to business during the 2014 enterprise survey, with almost a third (30.1 per cent) of firms identifying it as a major constraint. This chapter analyzes this obstacle in detail. 55 TABLE 14: CORRUPTION AND COURTS AS MAJOR CONSTRAINTS Corruption as Court as major FIGURE 27: FIRMS IDENTIFYING CORRUPTION AS A MAJOR major constraint CONSTRAINT (%) constraint 2008 2014 2008 2014 Tanzania 47.2 Courts 4.1 7.3 Corruption 12.8 30.1 Zimbabwe 38.3 Firm Size Malawi 30.1 Small 10.0 33.1 1.1 8.9 Zambia 29.8 Medium 15.1 30.5 6.3 5.4 Mozambique 25.4 Large 20.8 18.5 14.0 4.9 Kenya 21.3 Firm Type South Africa 16.9 Manufacturing 18.4 28.5 11.1 6.3 Sub−Saharan Africa 41.3 Service/retail 10.8 38.4 1.8 7.8 Location 10 20 30 40 50 Blantyre 29.4 5.1 Percent of firms identifying the courts system Lilongwe 30.9 10.3 as a major constraint Mangochi 15.2 0 Percent of firms identifying corruption as a major constraint Mzimba 39.6 19.3 Zomba 29.2 5.5 Source: Authors’ calculation based on data from WBES A third of firms in Malawi identified corruption as a major obstacle to enterprise development, reported mostly by largescale firms and those located in cities. 30.1 percent of enterprises that were surveyed in 2014 identified corruption as a major constraint obstacle, compared to 12.8 percent in 2008. 18.5 percent who identified this as a major obstacle were large-scale enterprises against 30.5 percent medium scale enterprises, and 33.1 percent small-scale enterprises. There were no significant differences in the perceptions between firms engaged in manufacturing and those in services sector. A significant percentage of firms in the major cities reported this as a huge constraint. Corruption is perceived to be a serious challenge in Malawi, although somewhat better than in some comparators countries. Corruption seems to be a widespread problem in the surrounding countries, with over a third of enterprises in most countries identifying it as a major constraint. Although the perception among enterprises of corruption as a major constraint is relatively better in Malawi (30.1 percent) than in the Sub-Saharan Africa region (41.3 percent), the country compares favorably only with Tanzania (47.2 percent) and Zimbabwe (38.3 percent). It compares worse than most of the comparator countries such as Zambia (29.8 percent), Mozambique (25.4 percent), Kenya (21.3 percent), and South Africa (16.9 percent). The court system in Malawi is reported to be a major constraint mostly for large enterprises. 7.3 percent of businesses identified the court system as a major constraint to business development, which is a deterioration from 4.1 percent in 2008. 14.0 percent of large scale enterprises identified this as a major constraint, against 6.3 percent for medium scale and only 1.1 percent for small-scale enterprises. Bribery 57 Bribery incidence30 appears to be increasing in Malawi. In the 2014 Enterprise Survey, Malawi firms reported experiencing high levels of bribery incidences, as the percentage of firms experiencing at least one bribe payment request within a year rose from 13.7 in 2008 to 24.2 percent in 2014. The average for Sub- Saharan countries was at 22.7 percent. Similarly, bribery is deepening31. Firms reported that in 20.2 percent of public transactions, a gift or an informal payment was requested in 2014, compared to 16.7 percent in Kenya and 15.7 percent in Tanzania. South Africa, Zambia and Zimbabwe reported 3.0 percent, 9.7 percent and 9.4 percent respectively. The depth of bribery in Malawi is below the average for Sub-Sahara (25.0 percent), although it worsened significantly compared to 8.5 per cent in 2008. TABLE 15: FIRMS IDENTIFYING CORRUPTION AS MAJOR CONSTRAINT 2008 2014 FIGURE 28: BRIBERY RATES Those identifying corruption as 12.8 30.1 major constraint 35 Those identifying court system 4.1 7.3 30 as major constraint Bribery Incidence 13.7 24.2 25 Manufacturing 9.9 16.6 20 Services 15.1 28.1 15 Small 14.5 25.7 Medium 12.6 27.2 10 Large 12.0 15.2 Bribery Depth 8.5 20.2 Manufacturing 5.8 15.8 to get water connection to get construction permit to get electric connection to get an operating license in meeting with tax officials to secure government contract to get import license Services 9.4 22.4 Small 8.6 21.4 Medium 8.1 24.2 Large 8.7 10.4 Firms expected to give gifts in meeting with tax officials 11.4 18.3 to secure government contract 21.7 33.0 to get an operating license 3.5 16.6 to get import license 0 26.7 to get construction permit 4.9 34.5 to get electric connection 12.6 25.3 2008 2014 to get water connection na 27.3 Source: Authors’ calculation based on data from WBES Firms reported facing challenges getting services such as permits and licenses in the course of doing their business as well as payment of taxes. Comparing results from the 2008 and 2014 enterprise surveys, there are growing proportions of firms expected to give gifts in order to meet tax officials, to secure government contracts, to get business operating licenses, and import licenses, to obtain construction permits, and to get connected to an electricity grid or water supply. The percentage of firms reporting these ranged between 16.6 and 34.5 percent (see table 14 below), up from as low as 0 to 12.8 percent in 2008. Striking examples of worsening situation between the two surveys is that 33 percent of firms interviewed 30 Bribery incidence refers to percentage of firms experiencing at least one bribe payment request 31 Bribery deepening refers to percentage of public transactions where a gift or informal payment was requested 58 in the 2014 survey (compared to 21.7 percent in 2008) reported that they were expected to give gifts to secure a government contract, and 34.5 percent of firms (compared to 4.9 percent in 2008) were expected to give gifts to get a construction permit. According to the 2014 Enterprise Survey, Malawi was one of the top three worst performers amongst comparable countries in the region with high corruption levels (in form of ‘gifts’) when enterprises get services from public institutions, as noted in the table below. TABLE 16: CORRUPTION LEVELS – COUNTRY COMPARISON A B C D E F G H I J Malawi 16.6 (3) 16.6 (2) 26.7(1) 34.5 (2) 25.3 (2) 27.3 (3) 20.2(1) 33.0 18.3(1) 24.2(2) Kenya 28.2 (1) 15.6 (3) 17.0(2) 34.6 (1) 17.6 (3) 30.9 (2) 16.7(2) 33.4(3) 17.4(2) 26.4(1) Tanzania 20.0 (2) 17.0 (1) 5.1 31.4(3) 25.3 (1) 20.4 15.7(3) 66.2(1) 14.6(3) 20.8(3) Mozambique 14.8 6.9 10.6(3) 3.9 14.6 16.2 9.7 32.1 9.8 12.4 Zambia 9.5 7.6 6.7 13.6 14.8 43.3 (1) 9.7 27.5 9.0 15.8 Zimbabwe 14.4 9.0 7.6 9.2 32.7 … 12.3 21.1 12.6 12.3 South Africa 15.1 0 2.7 0 6.7 4 3 34.2(2) 3.1 4.2 Sub-Sahara 25.9 16.1 16.8 25.9 22.5 23.3 17.4 33.8 17.2 22.7 Key Shaded cells = top three worst performing countries represented in the table A = Percent of firms expected to give gifts to public officials “to get things done” B = Percent of firms expected to give gifts to get an operating license C = Percent of firms expected to give gifts to get an import license D = Percent of firms expected to give gifts to get a construction permit E = Percent of firms expected to give gifts to get an electrical connection F = Percent of firms expected to give gifts to get a water connection G = % of public transactions where gifts or informal payment was requested H = % of firms expected to give gifts to secure government contract I = % of firms expected to give gifts in meetings with tax officials J = % of firms experiencing at least one bribe payment request Source: Authors’ calculation based on data from WBES 59 CHAPTER 6: OTHER RISK MEASURES This chapter looks at a number of other risk measures affecting private sector enterprises in Malawi. These include: taxation; practices of the informal sector; access to land; political instability; regulatory compliance; crime, theft and disorder; water and transportation. FIGURE 29: TOP TEN BIGGEST OBSTACLES TO ENTERPRISES IN MALAWI, 2008 AND 2014 Access to finance Electricity Corruption Tax rates Practices of the informal sector Access to land Political instability Business licensing and permits Crime, theft and disorder Transportation 0.0 20.0 40.0 60.0 Malawi 2014 Malawi 2008 Source: Authors’ calculation based on data from WBES 6.1 TAX RATES AND ADMINISTRATION In the 2014 enterprise survey, tax rates were listed as number four on the list of obstacles to enterprise development in the country. In response to a question of what managers consider as the biggest obstacle to their establishments, 10 percent indicated that it is tax rates. This compares unfavorably with the perception in 2008 survey when only 7.0 percent considered this as biggest obstacle. When asked to what extent this is an obstacle, 35.6 percent identified this as a major constraint compared to 15.6 percent in 2008. In the 2008 survey, most of the complaints about tax rates seem to have come from large scale enterprises in the services sector, but in 2014, it was mostly from medium and small-scale enterprises. 60 TABLE 17: TAX RATES AS A MAJOR CONSTRAINT ___________________________________ 2008 2014 FIGURE 30: LICENSING, TAXATION AND TAX ADMINISTRATION Firms identifying tax 15.6 35.6 rates as a major constraint 2008 2014 Firm Size Small 5.6 37.4 Percent of firms Medium 29.4 40.6 identifying business 9.0 Large 30.2 20.9 licensing and permits as a 11.2 Firm Type major constraint Manufacturing 23.9 36.4 Percent of firms Retail/services 12.7 40.5 identifying tax 9.0 Location administration as a major 21.1 Blantyre 34.4 constraint Lilongwe 37.6 Percent of firms 15.6 Mangochi 26.9 identifying tax rates as a Mzimba 33.5 35.6 Zomba 38.4 Number of visits with 2.6 2.0 tax officials 0.0 10.0 20.0 30.0 40.0 Source: Authors’ calculation based on data from WBES Most of the complaints about tax rates in 2014 were from the medium scale retail firms . 40.6 percent of medium scale firms in Malawi identified tax rates as a major constraint. This was followed by 37.4 percent of small firms, and 20.9 percent of large firms. 40.5 percent retail firms and 36.4 percent from manufacturing firms reported this as major constraint. In terms of location, firms that reported tax rates as an obstacle were mostly from the country’s major cities in Lilongwe, Blantyre and Zomba (compared to Mangochi and Mzimba). On the other hand, tax administration as an obstacle was cited by a significantly larger proportion of firms in 2014 (21.1 percent) compared to 2008 (9.0 percent) . Most of the complaints about this challenge came from the medium and smaller sized enterprises, and equally from the retail and the manufacturing sectors. It was identified as a major constraint by 24.1 percent of the medium enterprise firms, 22.6 percent of the small-scale firms and 11.4 percent of the large-scale firms. On average, 2 visits were required to hold meetings with tax officials in any given month in 2014, down from 2.6 visits in 2008. While this is better and below the average for Sub-Saharan countries at 2.2 visits, it is above the average for other comparable countries such as Tanzania, Mozambique and Zambia. There are a number of countries in SSA where average visits are much less. For example, in Mauritius, the average number of visits is reported at 0.5 days, Namibia (0.6 days), and Madagascar (0.8 days). Kenya managed to reduce the required number of visits from 6.7 days in 2007 to 1.5 days in 2013. 6.2 PRACTICES OF THE INFORMAL SECTOR Practices of competitors in the informal sector are ranked as the fifth biggest obstacle in the business environment in Malawi. The survey suggests that 90.3 percent of firms in the country formally registered their firms at the start of their operations, up from 78.6 percent in 2008. While this was the case, 71.7 percent reported competition with informal players as a challenge, down from 77.8 percent in 2008, but still higher than the Sub Saharan Africa average of 67.4 percent. It was reported as such across all categories, but was highest for manufacturing firms, followed by retail and other services. When asked to 61 what degree this is a constraint in the 2014 enterprise survey, 30.1 percent (against 26.4 percent in 2008) identified this as a major obstacle. 36.3 percent were small scale, followed by 27 percent from the medium scale firms and 13.2 percent from the large scale firms. 34.6 percent were non-exporters while 5.6 were exporters. TABLE 18: INFORMAL SECTOR COMPETITION AS MAJOR CONSTRAINT 2014 Firms Firms Number of Firms competing formally years firm identifying FIGURE 31: COMPETITION FROM INFORMAL against registered operated practices of unregistered when they without competitors in PLAYERS or informal started formal the informal firms (%) operations in registration sector as a 78.375.4 the country major 80.0 71.772.673.9 (%) constraint (%) 67.4 70.0 59.3 77.8 78.6 0.6 35.7 60.0 2008 48.549.6 50.0 38.5 45.3 45.047.4 2014 71.7 90.3 0.7 30.1 Firm Size 40.0 26.930.1 Small 70.7 87.9 0.3 36.3 Medium 75.2 92.7 1.7 27.0 20.0 11.3 Large 69.3 95.3 0.6 13.2 10.0 Firm Type 0.0 Manufacturing 76.5 89.7 0.6 29.9 Malawi Zimbabwe Kenya Tanzania Mozambique Zambia Sub−Sahara South Africa Service 69.2 90.7 0.8 30.1 Type Exporter 63.8 68.7 4.7 5.6 Non exporter 73.1 92.5 0.5 34.6 Ownership Male Top 72.6 89.4 0.8 27.9 Manager Female Top 68.3 95.7 0.1 43.9 Manager Source: Authors’ calculation based on data from WBES Practices of competitors in the informal sector seem to be a challenge across the comparable countries in the region. The percentage of those that identified this as a major constraint for Malawi was 30.1 percent, against 49.6 percent in Mozambique, 48.5 percent in Zambia, 47.0 percent in Zimbabwe and 47.4 percent in Tanzania. Kenya and South Africa have lower percentages than Malawi. The average for Sub- Saharan Africa was at 38.5 percent. 6.3 ACCESS TO LAND AND POLITICAL INSTABILITY In past years, the processes for acquiring land have not been investment-friendly, as they have been characterized by long, prohibitive and non-transparent processes. Easing access to land for industrial and commercial purposes is one of the key targets of recent land law reforms. The reforms in land governance have since commenced with the review and adoption of 10 land laws in 2016. Implementation of these new laws is expected to ease the process of acquiring land for commercial and industrial purposes through the facilitation of the Malawi Investment and Trade Centre. Acquiring land will also be facilitated by the establishment of an integrated land information system that is being developed with support from the World Bank. 62 Results from the 2008 and 2014 enterprise surveys indicate that access to land has been a growing challenge to private sector enterprises in Malawi. Only 1.7 percent of the firms reported that access to land is the biggest constraint that they faced in 2008, but this proportion rose to 7.1 percent in 2014. This is significantly higher than the average for Sub-Saharan Africa (4.7 percent) and neighboring countries Tanzania (5.1), and Mozambique (5.2). Access to land proved more of a constraint for firms with a female top manager than those with male top managers (see table 19 and Figure 32). TABLE 19: ACCESS TO LAND AND POLITICAL INSTABILITY Proportion of Proportion FIGURE 32: LAND AND POLITICAL INSTABILITY AS OBSTACLES - firms citing citing COUNTRY COMPARISON access to land political as biggest instability as obstacle (%) biggest 20 18.9 obstacle (%) 2008 2014 2008 2014 18 16 Access to Land 1.7 7.1 14 Political Instability 0.5 4.2 12 10.6 9.8 Firm Size 10 8.6 Small 0 6.1 0 4.1 8 Medium 3.8 10.2 0 4.2 5.2 5.1 Large 4.4 6.0 4.5 4.6 4.7 4.7 Firm Type 2.7 Manufacturing 6.4 6.0 0 3.1 1.1 1.1 Retail/services 0 9.3 0.7 1.0 0.5 0.6 0.3 Direct exporters 13.6 0 Sub−Saharan Africa Non exporters 0 3.9 Kenya Zambia Malawi Tanzania Zimbabwe Mozambique South Africa Location Blantyre 6.1 5.3 Lilongwe 7.9 2.8 Mangochi 19.7 0 Mzimba 20.7 0 Zomba 2.8 4.9 Female Top 11.8 0 5.5 manager Political instability Access to land Male Top manager 6.4 0.6 4.0 Source: Authors’ calculation based on data from WBES Concerns of political instability also appear to have increased in 2014 compared to the results from the survey in 2008. 4.2 percent cited this as the biggest obstacle according to results from the 2014 enterprise survey compared to only 0.5 percent in the 2008 survey. Evidence suggests that political instability constraints were high in the Sub-Saharan African countries including some of the comparable countries such as Zimbabwe and Kenya. It was certainly more of a concern for enterprises in Malawi than it was for those in neighboring countries of Zambia, Tanzania, South Africa and Mozambique (Figure 32) 6.4 BUSINESS LICENSING AND PERMITS Challenges associated with obtaining business licensing and permits are ranked as the eighth biggest obstacle to enterprises in Malawi. 3.6 percent of the respondents to the 2014 enterprise survey identified this as an obstacle (no change compared to perceptions in 2008). When asked to what degree this is a 63 challenge 11.2 percent identified this as a major obstacle. A bigger percentage was from the small scale enterprise compared to medium or largescale enterprises and from firms with female top managers. Government bureaucracy is another challenge cited by private sector. Firms in the survey reported that 5.0 percent of senior management’s time is spent dealing with fulfilling requirements of government regulations. This is better than the average for sub-Saharan countries and for countries such as Kenya, South Africa and Zambia. It is however, still very high, and worsened compared to 2008. TABLE 20: BUSINESS LICENSING AND PERMIT 2014 Percent of firms Senior identifying management FIGURE 33: FIRMS IDENTIFYING BUSINESS LICENSING business time spent AND PERMITS AS MAJOR CONSTRAINTS (%) licensing and dealing with permits as a the major constraint requirements (11.2%) of government 34.2 regulation 35.0 (5.0%) 30.0 Firm Size 25.0 17.8 17.4 18.7 Small 13.0 4.0 20.0 13.7 Medium 9.1 6.8 15.0 9.4 11.2 Large 8.2 5.8 10.0 3.0 Firm Type 5.0 Manufacturing 11.2 6.0 0.0 Zambia Sub−Saharan Africa South Africa Kenya Malawi Tanzania Zimbabwe Mozambique Service/retail 10.8 4.2 Type Exporter 13.0 5.6 Non exporter 11.8 5.1 Male Top 11.2 5.2 Manager Female Top 7.6 3.9 Manager Source: Authors’ calculation based on data from WBES While being within the average levels for Sub-Saharan countries, significant amounts of time resources are spent to obtain necessary documentation to do business in Malawi . Evidence from the 2014 Enterprise Survey suggests that it takes 19 days on average to obtain an operating business license in Malawi. This compares favorably with Mozambique, South Africa, and Zambia and is similar to the average for Sub-Saharan countries, but worse than countries such as Kenya, Tanzania and Zimbabwe. Malawi firms reported 36.2 days to obtain construction related permits, one of the best in the region and much better than the average for Sub-Saharan countries. Except for Mozambique, Malawi records the least number of days to obtain an import license. 64 FIGURE 34: NUMBER OF DAYS TO GET LICENSES AND PERMITS 192.2 150.0 38.6 41.7 55.0 50.6 36.2 41.3 41.3 50.0 22.7 19.0 22.0 35.2 18.8 8.1 0.0 Sub−Saharan Africa Mozambique Kenya Malawi Zambia Zimbabwe Tanzania South Africa Days to obtain a construction−related permit Days to obtain an operating license Days to obtain an import license Source: Authors’ calculation based on data from WBES 6.5 CRIME, THEFT AND DISORDER Though still high, there was a slight improvement in the perception by firms in Malawi on the extent to which crime, theft and disorder was a constraint in 2014 in comparison with the 2008 findings. The proportion of firms that identified crime, theft and disorder as a major constraint in the country was 20.7 percent in 2014 a slight improvement from 22.8 percent in 2008. In terms of size of firms, 20.0 percent of small scale enterprises and 29.1 percent of medium scale enterprises reported this as a major constraint compared to 9.2 percent of large scale enterprises. A higher percentage of retail firms also reported this compared to those engaged in manufacturing. Crime, theft and disorder was reported by 38 percent of the enterprises in South Africa, 33.6 percent in Mozambique, 21.2 percent in Kenya, 21.1 percent in Tanzania, 10.5 percent in Zambia, and 6.6 percent in Zimbabwe. The Sub-Sahara average was 17.1 percent. 65 TABLE 21: LOSSES DUE TO THEFT, CRIME, VANDALISM FIGURE 35: CRIME, THEFT AND DISORDER - Firms Securit Firms Losses Products Percent of COUNTRY COMPARISON payin y costs experien due to shipped firms g for (% of cing theft to supply identifyin securit annual losses and domestic g crime, y (%) sales) due to vandalis markets theft and 80.0 70.0 theft m that were disorder 60.0 and against lost due as a major vandalis the firm to theft constraint 40.0 m (%) (% of (% of 30.0 20.0 annual product 10.0 sales) value)* 0.0 Zimbabwe Malawi Mozambique Zambia Tanzania Kenya Sub−Saharan Africa South Africa 2008 90.4 6.6 50.1 11.7 22.8 2014 70.1 3.2 33.5 2.2 1.5 20.7 Firm Size Small 59.1 2.5 29.6 2.0 1.9 20.0 Medium 81.7 3.8 36.9 3.0 1.2 29.1 Large 90.2 4.3 41.6 1.8 1.1 9.2 Business Sector Manufactur 67.3 3.6 34.3 2.4 17.2 ing Retail 65.1 2.3 28.2 1.7 24.6 Other 75.2 3.3 36.0 2.4 21.2 service Source: Authors’ calculation based on data from WBES Compared to 2008, the percentage of firms experiencing losses due to theft and vandalism also reduced in 2014. 33.5 percent of the firms reported experiencing revenue losses due to this challenge, down from 50.1 percent in 2008. The large and the medium scale enterprises and the manufacturing firms experienced heavier losses compared to the small-scale enterprises, as well as those engaged in the retail services. To cope with theft challenges and related losses, the enterprises reported getting services of private security firms. 70 percent of firms reported paying for security, again, an improvement over 2008 when 90.4 percent spent money for services of private security firms due to these challenges. Nonetheless, theft and vandalism cause significant losses to enterprises in Malawi . 1.5 percent of product value being shipped to supply domestic markets is normally lost due to theft according to the 2014 survey. Malawi and Zambia firms for instance reported highest losses due to theft and vandalism at 2.2 percent of annual sales and 2.3 percent of annual sales, respectively. These two countries also registered highest security cost (as percentage of annual sales), at 4.9 percent and 6.9 percent, respectively. The average for Sub-Saharan Africa is 6.5 percent. In general, theft and vandalism levels in Malawi improved significantly from 2008, but do not compare favorably with other countries in the region. More foreign firms pay for security services compared to the domestic firms, costing 5.5 percent, compared to 4.6 percent of their annual sales. Consequently, foreign firms experienced lesser losses than domestic firms at 6.1 percent versus 9.1 percent, respectively. 0.9 percent loss of products shipped to domestic markets were reported by foreign firms, compared to 1.9 percent by domestic firms. Overall, there is a significant difference in the extent to which crime, theft and disorder is experienced as a challenge, with only 8.5 percent of foreign firms citing it as a major constraint, compared to 22.5 percent of domestic firms. 66 Trucks awaiting entry at the Dedza Boeder 6.6 TRANSPORT AND LOGISTICS The transport sector in Malawi faces a number of challenges. Producers in the country face high transport costs in sourcing raw materials and in delivering outputs to the domestic, regional, and global markets. Road transport accounts for 70 percent of all freight, and 99 percent of all passengers. However, the overall quality of the road network, particularly at the secondary and tertiary level, is poor. The rail sector is also not fully functional. It was also noted that the road and rail infrastructure were particularly in bad shape around 2008. TABLE 22: TRANSPORT CHALLENGES 2008 2014 FIGURE 36: TRANSPORT AS A MAJOR CONSTRAINT -COUNTRY Firms identifying transportation COMPARISON 11.4 2.7 as biggest obstacle Firms identifying transport as a 24.6 15.7 major constraint (%) 36.7 Small scale 22.9 17.2 40.0 Medium scale 25.7 13.7 35.0 Large scale 30.3 13.5 30.0 26.1 Manufacturing 34.8 16.6 25.0 21.6 23.0 Retail n.a 18.5 20.0 15.7 16.4 17.1 Services 21.1 15.2 15.0 Exporter (portion of sales are exports) 34.2 15.3 10.0 3.9 Domestic only firm (no exports 24.4 15.7 5.0 Top Manager, Female 30.9 11.5 0.0 Malaw i Sub−Saharan Africa South Afric a Zambia Kenya Zimbabwe Tanzania Mozambique Top Manager, Male 23.4 15.9 Firm has no foreign ownership 24.4 17.0 Firm has at least 10% foreign 25.4 12.9 ownership Lilongwe 29.0 Blantyre 25.8 Mangochi 9.2 Mzimba 22.4 Source: Authors’ calculation based on data from WBES According to the results of the enterprise surveys, there has been some improvement in 2014 in the perception of business persons over transport challenges compared to 2008, but none-the-less, this remains a significant challenge to business enterprises. This obstacle ranked 10th among the biggest obstacles to enterprise development in the 2014 enterprise survey. It was ranked 2 nd biggest obstacle, after access to finance, in the 2008 enterprise survey, but has been replaced in this position by electricity constraints. This points at some improvements that have happened in between, but could also point at the increase in electricity constraints relative to those related to transport in the operations of firms. In the 2014 survey, only 2.7 percent of top managers and enterprise owners cited this as the top most obstacle, an improvement from 11.4 percent that reported the same in 2008. This was slightly above the average for Sub-Saharan Africa. When asked to what extent transportation is a challenge, 15.7 percent of the firms identified this as a major constraint, again, an improvement from the 2008 perception when 24.6 percent of firms identified this as a major problem. While this is lower than other comparator countries in the region, the impact of transport and logistics challenges for Malawi are more pronounced, considering that Malawi is a landlocked country. The small-scale enterprises seem to be more concerned by this challenge, compared to the medium and the large-scale enterprises, more especially in the retail subsector. 68 6.7 WATER SHORTAGES Earlier on in Chapter 4, it was highlighted that Malawi faces challenges of water unavailability to industry and agriculture production. Some of the key issues that were identified as causes are: (i) increasing climatic and hydrological variability, and limited resilience to floods and droughts; (ii) limited stock of water storage and irrigation infrastructure; (iii) degradation of watersheds leading to increased soil erosion and sedimentation; (iv) inefficient water utilities that lack autonomy and accountability; and (v) limited maintenance of distribution infrastructure resulting in significantly high leakage rates. It was also noted that the supply of water in the country is growing slower than the demands of a growing population. The 2014 enterprise survey confirmed that water shortages and disruption pose challenges to Malawi enterprises. In the 2014 survey, firms reported experiencing on average 5.3 insufficiencies of water in a typical month, a concerning deterioration from 0.5 in 2008. This is the highest estimate amongst comparable countries in the region, and almost triple the average number of days of water disruptions in Sub-Saharan African countries (see figure 37). Water outages were reported more prominent by enterprises in Blantyre city at 6.4 insufficiencies in a typical month, compared to 4.1 in Lilongwe. Large and small scale enterprises identified themselves with this problem more than the medium enterprises, with more effects noted on manufacturing firms. TABLE 23: WATER INSUFFICIENCIES FIGURE 37: WATER INSUFFICIENCIES IN A TYPICAL MONTH Malawi 2008 2014 Number of Water 0.5 5.3 Insufficiencies in a typical month Firm Size Small 0.5 5.6 Medium 0.4 3.9 Large 0.5 5.8 Firm Type Sub-Saharan… Manufacturing 0.5 5.3 Kenya Zambia Malawi Zimbabwe Tanzania Mozambique South Africa Retail/services 0.5 n.a. Location Blantyre 6.4 Lilongwe 4.1 Source: Authors’ calculation based on data from WBES 69 CHAPTER 7: CONCLUSIONS FOR POLICY ON INVESTMENT CLIMATE IN ORDER TO IMPROVE PRODUCTIVITY The foregoing chapters have analyzed the constraints that firms in Malawi’s current investment climate face. A few issues have come out requiring policy interventions that can improve the investment climate and enhance the productivity of firms in the country. From the analysis it is clear that policymakers face one distinct pathway, namely, the reinforcement in the approach to private sector development in which case, given all the challenges the country faces, the pace of reform for Malawi needs to be focused on critical areas, be more rapid, deep and consistent. This is necessary if Malawi is to enhance and maintain its competitiveness in a fast moving world. This paper proposes that the focus should be in four main areas as follows: (a) Increase the productivity of firms, particularly medium sized ones: Advantage must be taken of Malawi’s higher levels of labor productivity and lower unit cost levels than neighboring countries to attract investment into labor-intensive industries. While doing this, labor productivity of medium firms needs to be improved, as does their use of capital. An increase in labor-intensive industries can create jobs that through increasing incomes, can positively reinforce capital utilization for domestic markets. Market access to regional and global markets can also enhance capital utilization as more products are produced for those markets. Further to these partial productivity measures, total factor productivity needs to be improved inter alia through development of an innovation ecosystem and frameworks that facilitate high levels of technology diffusion; and research and development, education and skills development that promote agricultural production and feed into the labor needs of emerging manufacturing sub-sectors. (b) Improve access to finance: Access to finance was the biggest obstacle to enterprise development in 2014 and skewed in disfavor of poor local and rural enterprises. A typical firm owned by local Malawian entrepreneurs in rural areas, especially from the firms headed by women, and those that are involved in production for the domestic market seemed to be more concerned about this challenge. The proportion of loans requiring collateral and the average value of collateral requirements are very high in Malawi. As a result, only a small proportion of working capital in Malawi is financed by banks, and this is skewed in favor of large scale enterprises, those engaged in the manufacturing, those engaged in export trade and the firms that are foreign owned. Most of the firms depend heavily on funding their working capital and investments through own funds. From 201432, there have been positive developments aimed at improving access to finance. For example, the Credit Reference Bureau Act was amended in 2016, and the Collateral Registry was launched in the same year, opening up prospects for responsible reduced risk in lending and movable-asset lending respectively. Implementation of the two reforms is in its early stages but shows some promising results. A Warehouse Receipts Bill has been drafted, and is scheduled for submission to parliament soon. The payments system infrastructure continues to be developed and other innovations in electronic payments are mushrooming. Thoughts around the setting up of institutions for long term development financing are featuring in the dialogue on several fora including the Public Private Dialogue (PPD) meetings. Government seems to be committed to this, but implementation of these plans needs to be expedited. On a broader level, efforts on macroeconomic fundamentals are bearing fruit, with the Reserve Bank of Malawi’s policy rate declining to 22 per cent in 2017 and inflation reducing to 14.2 percent in May 2017. 32 Establishment of Export Development Fund (EDF) is one of the earlier initiatives to increase access to finance by firms engaging in export. 70 Concerted efforts need to be engaged to ensure the following: (i) That the macro economy continues to stabilize, interest rates lower, and sufficient space is created for the private sector to thrive including through settlement of arrears to ensure private sector players are able to finance their operational costs and implement expansion plans. (ii) That the improvement of credit reporting infrastructure, collateral registry, warehouse receipts as well as commodity exchange frameworks are fully developed and implemented and that ensure that these respective systems are functioning effectively. (iii) That the country develops a long term financing institution to provide concessional lending to the private sector and hence ensure sustainable expansion of their operations. (iv) That new financial instruments are introduced, especially those targeting the disadvantaged. Such efforts need to ensure increased access to finance by small and medium-scale enterprises with attention to the disadvantaged identified across the different categories. In particular, the gender disparities noted in access to finance need to be interrogated further and addressed. Examples include leasing, and factoring, which can be done within the framework of the Collateral Registry. (v) To support the missing middle, effective channeling of rapidly accumulating pension funds needs to be given priority including through improvements to the Malawi Stock Exchange so that more companies are publicly listed. It is recommended that IFC should strengthen its support on the access to finance agenda, such as to provide funds for agriculture financing and for FIs in Malawi to develop new and innovative lending instruments such as for specific investments or specific industries. The Financial Sector Technical Assistance Project financed from the World Bank is due to close in August, 2017, but a number of activities on that project are yet to be concluded. Some of the critical elements include digital solutions to strengthen micro finance institutions and SACCOs (including those in remote rural areas) and integration of mobile network operators to the national switch infrastructure. It is recommended that these activities be seen to their conclusion. RBM is developing a leasing framework together with a diagnostic study examining the applicability of a long term finance guarantee scheme. It is recommended that this work be expedited along with efforts to establish long term financing institutions, such as a development bank in Malawi. The Export Development Fund operations need to be improved and that the institution should be made more visible and able to reach out to more clients that require financing for their enterprise operation. IFC could be engaged to support these areas. There is need to review the credit culture in Malawi to facilitate the flow of credit. In theory, this could be a result of low levels of financial literacy, limited ability to manage collateralized loan portfolio by financial institutions, and lack of consequences as a result of limited enforcement action. Development partners including the World Bank should support Government to deal with these issues, including the implementation of robust financial literacy programs, public awareness of various opportunities, reviewing legal framework to protect FI in cases of default, identification of necessary adjustments to the legal and regulatory environment and development of disclosure frameworks for banks and MFIs products. 71 The Malawi Resilience Development Policy Financing (DPO) has so far helped to support Government efforts to implement good policies which promote favorable macroeconomic environment, that in turn translate into conducive business environment. While asking government to ensure that there should be long lasting progress in promoting macroeconomic stability, follow up World Bank support to the DPO should aim at strengthening this aspect, which ultimately will lead to a much more favorable business environment through sustained low inflation and interest rates. (c) Focusing finite public sector resources where they can achieve the greatest impact: Improvements in Infrastructure: Malawi needs to focus finite public sector resources where they can achieve the greatest impact. The evidence of this analysis points to the poor supply of utilities, particularly electricity and water, as major binding constraint to growth. In the 2014 enterprise survey, 24.8 percent of firms identified electricity as a major obstacle to their businesses. This was the second biggest obstacle overall, but was first for largescale enterprises in the manufacturing subsector. Firms expressed concerns about challenges experienced to get connected to the electricity grid. Infrastructure improvements in this area and transport, therefore, should be prioritized, to ensure that enterprises have the basic requirements to operate efficiently. It is imperative that the legal and policy reforms that have commenced in the energy sector, are effectively implemented to lead to the sound regulation of the sector that will translate to substantial investment. Within this, it is essential that independent power producers be effectively brought on stream utilizing all alternative sources of renewable energy to ensure adequate supply of electricity in both rural and urban areas. The national grid needs to continue to be expanded to uptake and efficiently transmit the resultant increased generation of electricity. Efforts to implement interconnectors with Malawi’s neighbors such as Mozambique and Zambia must also be reinforced to ensure that Malawi participates in electricity trade through the Southern African Power Pool (SAPP). These efforts should go along deeper efforts to improve the governance and efficiency of utility suppliers Alongside the upgrading of electricity, the government should also prioritize investments into water and transport infrastructure. On transportation, the Government should prioritize efforts to develop the Nacala Corridor, and improve railway connectivity as well as road networks in order to reduce domestic and international trade costs and make Malawi’s products more competitive. (d) Address policy and regulatory challenges to increase predictability and transparency and curb corruption. Corruption is increasingly seen as a big challenge to business activities in Malawi. This is experienced more significantly in the areas where businesses interact with Government (G2B services) which are also key factors for successful setting up and operations of business enterprises, such as on getting permits for construction, getting electricity connected, obtaining import licenses and acquiring land. Alongside corruption, Malawi faces challenges of lack of predictability in implementation of policy measures, which would otherwise encourage and facilitate businesses. Measures need to be taken, especially in Government institutions, to ensure transparency and accountability, as well as to improve the provision of services that facilitate businesses. In this way, private sector would not lose resources and time to get services that support growth of their enterprises. Strong measures are required to curb corruption in all its forms. A number of measures are proposed as follows: - (i) Ensure simplification and automation of G2B systems such as licensing and business permit systems. With time there is also a need to automate the majority of payment transactions with Government institutions such as for taxes. These would play a big part 72 in curbing corruption in such areas of interaction between the private sector and Government. This should be preceded by building and implementing necessary literacy programs. Lack of clarity is a major enabler of corruption and demands of informal payments by government officials. Much greater efforts are needed to make existing regulations, including tax and licensing requirements, simpler, more accessible and easier to implement. (ii) Malawi should ensure highly sound public financial management system and processes. Improving the IFMIS infrastructure, and the reductions of cash transactions in favor of e- payments for Government services are some of the important steps that Government should expedite on. (iii) The Anti-Corruption Bureau and other governance institutions should be adequately financed and be provided with the necessary independence to effectively execute their mandates. Implementation of reforms in land governance (new set of land laws) that have commenced should be effectively done to ensure predictability and ease in acquiring land by investors. Automation of land registration should be prioritized to ensure improvement on transparency. Government institutions mandated to provide land to investors, such as the Malawi Investment and Trade Centre, should be empowered to undertake their functions effectively. Some of the old protective laws such as the Control of Goods Act and the Special Crops Act should be reviewed to allow for transparency and predictability in investment decisions and trade across the country’s borders particularly for agriculture and agribusiness. Deeper efforts will be required to both ensure adequate consultation in policy changes, build in mechanisms that prevent ad hoc policy shifts, and address inconsistencies between the practices of different regulatory bodies. Apart from the automation of tax processes highlighted above, a comprehensive tax review needs to be prioritized, with due regard to feasibility assessments of spatial zones, which could potentially provide special tax incentives as an inherent feature of their design. 73 BIBLIOGRAPHY Clarke G. R. G. (2010). Are Managers' Perceptions of Constraints to Growth Reliable? Evidence from a Natural Experiment in South Africa. Texas: A&M International University. Clarke, G. R. G. (2011). Assessing How the Investment Climate Affects Firm Performance in Africa: Evidence from the World Bank’s Enterprise Surveys, World Background paper IV, Washington DC: World Bank. Clarke, G. R. G. (2012). Manufacturing Firms in Africa: Some Stylized Facts about Wages and Productivity, Texas: A&M International University. Clarke, G. R. G. (2014). Firm Performance in the West Bank and Gaza. Journal of International Finance Studies 14(3). Clarke, G. R. G. (2015). The Impact of Different Aspects of the Investment Climate on Total FactorProductivity: Evidence from the World Bank's Enterprise Surveys. Texas: A&M International University. Clarke, G. R. G., Qiang, C. Z., and Xu, L. C. (2015). The internet as a general-purpose technology: Firm- level evidence from around the world. Economics Letters. Dollar, D., Hallward-Driemeier, M., and Mengistae, T. (2005). Investment climate and firm performance in developing countries. Economic Development and Cultural Change 54(1). Escribano, A., and Guasch, J. L. (2005) Assessing the impact of the investment climate on productivity using firm-level data: Methodology and the cases of Guatemala, Honduras and Nicaragua . Policy Research Working Paper No. 3621.Washington DC: World Bank. Escribano, A., Guasch, J. L., Pena, J., and de Orte, M. (2005) Investment climate assessment on productivity and wages: Analysis based on firm level data from selected South East Asian countries. Washington DC: The World Bank. Felipe, J. (2008). What Policy Makers Should Know About Total Factor Productivity. Malaysian Journal of Economic Studies, Vol. 45 No. 1, 2008. Fisman, R., and Svensson, J. (2007) Are corruption and taxes really harmful to growth? Firm level evidence. Journal of Development Economics 83(1). Gatti, R., and Love, I. (2008) Does access to credit improve productivity? Evidence from Bulgaria. Economics of Transition 16(3): 445-465. Gonzelez, A., Lopez-Gordova, J.E., Valladares, E.E. (2007). The Incidence of Graft on Developing Country Firms, World Bank Policy Research Working Paper, Washington DC: World Bank. Government of Malawi (various years), Annual Economic Reports, Lilongwe, Ministry of Finance, Economic Planning and Development Government of Malawi. (2011). Malawi Growth and Development Strategy II. Lilongwe: Ministry of Finance, Economic Planning and Development 74 Greene, W. (2002) Econometric analysis. 5th edition. New Jersey: Prentice-Hall. Halvorsen, R., and Palmquist, R. (1980). The interpretation of dummy variables in semilogarithmic equations. American Economic Review. Harrison, A. E., Lin, J. Y., and Xu, L. C. (2014). Explaining africa's (dis)advantage. World Development 63(1): 59-77. Iarossi G., and Clarke, G. (2011). Nigeria 2011 : An Assessment of the Investment Climate in 26 States, Washington D.C., World Bank Kaplan, D. S., and Pathania, V. (2010). What influences firms' perceptions? Journal of Comparative Economics 38(4): 419-431. Kumbhakar, S. C., and Lovell, C. A. K. (2000). Stochastic frontier analysis. Cambridge: Cambridge University Press. Levinsohn, J. (2008). Comments on methodology used by escribano, guasch, and co-authors in analyzing World Bank investment climate surveys. Ann Arbor: University of Michigan. Levinsohn, J., and Petrin, A. (2003). Estimating production functions using inputs to control for unobservables. Review of Economic Studies 70(2): 317-341. Malomo, F. (2013). Factors influencing the propensity to bribe and size of bribe payments: Evidence from formal manufacturing firms in west Africa. Brighton: University of Sussex. Malawi Confederation of Chambers of Commerce and Industry (2015), Malawi Business Climate Survey. Blantyre: MCCCI Malawi Confederation of Chambers of Commerce and Industry (2016), Malawi Business Climate Survey. Blantyre: MCCCI Olley, S., and Pakes, A. (1996) The dynamics of productivity in the telecommunications equipment industry. Econometrica 64(6): 1263-1298. Pakes, A. (2008) Comments on methodology used in analyzing World Bank investment climate survey. Cambridge: Harvard University. Record, R., Clarke, G.H., Chilima, E. (2016). Malawi Undersized Private Sector, What are the Constraint to Higher productivity and Increased Competitiveness? Malawi Country Economic Memorandum 2016 Background Paper Series. Reserve Bank of Malawi. (2014a), Financial Institutions Supervison, Annual Report 2014. Lilongwe: Reserve Bank of Malawi. Reserve Bank of Malawi. (2014b), Financial Stability Report June 2014. Lilongwe: Reserve Bank of Malawi. Reserve Bank of Malawi (2016), Financial Stability Report December 2016, Lilongwe: Reserve Bank of Malawi. Deraniyagala, S. and Kaluwa, B. (2011) Macroeconomic policy for employment creation: The case of Malawi, ILO Employment Working paper. Geneva: International Labor Organization. 75 Svensson, J. (2003). Who must pay bribes and how much? Evidence from a cross section of firms. Quarterly Journal of Economics 118(1): 207-230. Tybout, J. R. (2003). Plant -and firm- level evidence on 'new' trade theories. In E. K. Choi & J. Harrigan eds., Handbook of international trade (pp. 389-415). Malden, Mass.: Blackwell Publishers. Van Ark, B. (2014). Total Factor Productivity: Lessons from the past and directions for the future. [Online]. Available at: https://www.nbb.be/doc/ts/publications/wp/wp271en.pdf Verhoogen, E. (2008). Comments on escribano-guasch methodology for estimating the effects of the investment climate on productivity. New York: Columbia University. World Bank. 2007. Turkey - Investment climate assessment: from crisis to private sector led growth. Washington D.C.: World Bank. World Bank. (2006). Malawi Investment Climate Assessment. Washington D.C.: World Bank. World Bank. (2014). Lao PDR Investment Climate Assessment: Policy Uncertainty in the Midst of Natural Resource Boom. Washington D.C.: The World Bank. World Bank. (2016a). An Assessment of the Investment Climate in Nigeria, The Challenges of Nigeria’s Private Sector, Washington, D.C.: World Bank. World Bank. (2017a). Malawi Economic Monitor – Harnessing the Urban Economy, Washignton D.C.: World Bank World Bank. (2017a). Doing Business Indicators. Washington, D.C.: World Bank. World Bank. (2017b). World development indicators. Washington, D.C.: World Bank World Bank. (2017c). Enterprise Survey Online Database. Washington D.C.: World Bank Group Xu, L. C. (2011). The effects of business environments on development: Surveying new firm-level evidence. World Bank Research Observer 26(2): 310-340. 76 ANNEX 1: MEASURING FIRM PERFORMANCE USING ENTERPRISE SURVEY DATA The analysis presented in this background paper focuses on several measures of firm productivity. These are calculated in a uniform way in all countries with available Enterprise Survey data from between 2006 and 2015.The accounting data is generally lagged one year from the year of the survey. So, for example, the accounting data for a 2015 survey will be for fiscal year 2014. The World Bank Enterprise Survey collect financial data in the local currency in the country being surveyed. To compare firm performance across countries, there is need to convert financial data into a common currency in a single year (i.e., to control for inflation and exchange rate differences). For surveys conducted between 2006 and 2010, data are inflated to 2010 values in local currency using GDP deflator33. For surveys conducted after 2011, the values are deflated for a different year. The values are then converted into US$ using 2010 exchange rates34. However, it is important to note that since most firms in the sample sell their products primarily in local markets, exchange rates have to be close to their equilibrium values in 2010 for these comparisons to be very accurate. If the exchange rate in a given country is over- or under-valued, the comparisons will under- or overstate firm performance for that country. Note that this proviso only applies to performance measures that are measured in U.S. dollars. For measures that are ratios such as capital productivity or unit labor costs, the exchange rates will cancel out during the calculations. As a result, these measures are mostly unaffected by exchange rate fluctuations. The individual measures are constructed in the following way; Value-added. Value-added is value of the goods and services that the firm produces less the cost of the raw materials (such as iron or wood) and intermediate inputs (such as engine parts or textiles) used to produce the output. Output is measured in local currency not in physical units. The cost of raw materials, intermediate inputs, electricity and fuel are subtracted from output to get value-added. Firms report electricity and fuel costs separately from raw materials and intermediate inputs. Firms that do not report sales or raw materials and intermediate inputs are dropped from the analysis. Electricity and fuel costs are treated as if they are zero for firms that do not report electricity or fuel costs (i.e., the firms are not dropped).This is done because dropping firms that do not report electricity or fuel costs would have a significant impact on sample size and because electricity and fuel costs are small relative to sales and raw materials. Number of workers. The number of workers is the number of permanent and temporary full-time workers. Temporary workers are weighted by the average length of employment for these workers. So, for example, if the average length of employment for a temporary worker was 6 months, the weight for temporary workers would be ½. If this is not done, it would result in very small sample estimates. Data on part-time workers is not collected in most countries (outside of Sub-Saharan Africa) and so part-time workers are omitted for reasons of comparability. In practice, for countries with data on part-time workers, including these workers do not have a large effect on relative rankings. Firms that do not report permanent or temporary workers are dropped for measures that use workers (e.g., value-added per worker). 33 GDP deflators is used rather than sector specific deflators because sector specific deflators are not available for most countries. 34 The 2010 values are deflated using a common year’s exchange rate because of concerns that fluctuations in the value of the dollar might otherwise make comparisons difficult. 77 Labor Productivity. Value-added per worker is the basic measure of labor productivity used in this paper. It is value-added divided by the number of full-time workers in the firm (see above). Firms that produce more output with less raw material and fewer workers have higher labor productivity. Capital Intensity. There are two measures of capital in the Enterprise Survey. The first measure is the book value of capital. For firms that keep detailed financial accounts, this measure should be the value of capital taken from those accounts. For other firms, it will either be omitted or estimated by the manager. In the World Bank’s survey implementation manual, this variable is described as follows: “The net book value represents the actual cost of assets at the time they were acquired, including all costs incurred in making the assets usable (such as transportation and installation) minus depreciation accumulated since the date of purchase.” The second measure is the sales value of capital. The manager is asked to estimate the value of the capital if sold in its current condition. Although this is probably closer to the true value of the capital, it has some shortcomings. In particular, when markets for capital equipment are thin, it might be difficult for the manager to give an accurate estimate. The implementation manual notes: “Ask the manager to estimate the market value if all of the equipment, land and buildings were sold on the open market. If the respondent states that there is no market, ask how much the respondent would be willing to pay for the capital, knowing what it can produce in its current condition. Estimate how much it would cost to buy machinery in the current market which is similar in terms of age and characteristics.” In the empirical analysis, focus is placed on the sales value of capital. This is done because it is closer to the economic concept of capital. Capital intensity is capital per worker. This is calculated by dividing capital by the number of workers. Firms that do not report these measures have to be dropped when calculating total factor productivity. Capital productivity. This is the ratio of value-added to capital. It can be constructed either using the book value or sales value of capital. When a firm produces a lot of output with little capital, capital productivity will be high. Firms that do not provide information on capital or enough information to calculate value- added are dropped. Total factor productivity/technical efficiency. This measure of productivity takes both labor and capital use into account. The methodology is described in detail in Annex 3. Labor costs per worker. The cost of labor is the cost of wages, salaries, bonuses, other benefits, and social payments for workers at the firm divided by the number of workers. The data is taken from the firms’ accounts. It includes wages and salaries paid to all workers and managers – not just production workers. We divide this by the number of workers to get labor costs per worker. Firms are only dropped from these averages when they do not report labor costs or workers. Unit labor costs. This measure is labor costs divided by value-added. Although it is an approximation to true unit labor costs (i.e., it measures output in dollars rather than as physical measure of production), it can be calculated using information from the Enterprise Surveys. Unit labor costs are higher when higher labor costs are not fully reflected in higher productivity. Employment and sales growth. Firms are asked about employment in the most recent fiscal year and two years earlier (e.g., for 2015 surveys, firms report 2013 and 2014 employment). Because the question for lagged employment only asks about full-time permanent employees, we use this as the measure of employment. That is, we do not include temporary or part-time workers for either the initial or final year. 78  The formula that we use is:  Employment Growth    Employmentt  Employmentt 2  1   (1) 2 Rather than using the initial value of employment as the denominator when calculating the growth rate, we use the approach used in Davis and Haltiwanger (1992). This measure is preferable to a measure that divides by initial employment because it does not become very large when initial sales or employment are small. We divide it by the growth rate by two to get an annual growth rate. It is important to note that initial employment is reported retrospectively. So, for example, for a 2015 survey, the manager reports employment in both 2013 and 2014 during the same 2015 interview. That is, this calculation is based on recall data reported in 2015 not on data from separate surveys in 2013 and 2014. We use an analogous measure for sales growth. Sales in both years are deflated into 2010 local currency units to get real, rather than nominal sales growth. See the description above for information on deflators. 79 ANNEX 2: CROSS COUNTRY COMPARISON USING ENTERPRISE SURVEY DATA In this annex, graphs are presented that plot each of the partial productivity measures discussed in the main body of the report and in annex 1 for countries where Enterprise Surveys was conducted, against per capita GDP35. The regression line comes from a simple linear regression of the partial productivity measure on per capita income and a constant. Within countries, the partial productivity measures are weighted medians or means. The regressions are based on unweighted country-level regressions. This is done to avoid giving large countries greater weight in the regressions. Although the graph only includes countries with per capita GDP less than US$ 12,000, these countries are excluded only for presentational purposes—they are included when calculating the regression line. If a point is above the regression line, the partial productivity measure is greater than would be expected given the county’s per capita income. If it is below the regression line, the opposite is true. FIGURE 38: MALAWI LABOR PRODUCTIVITY - HIGHER THAN WOULD BE FIGURE 39: MALAWI CAPITAL INTENSITY - HIGHER THAN WOULD BE EXPECTED GIVEN THE LEVEL OF DEVELOPMENT CONSIDERING THE COUNTRY'S LEVEL OF DEVELOPMENT $22,500 $20,000 Value-added per worker $20,000 $17,500 $17,500 Capital per worker (sales value, $15,000 $15,000 (2009 US$) $12,500 $12,500 $10,000 $10,000 $7,500 $5,000 2009 US$) $7,500 $2,500 $5,000 $0 $2,500 $0 $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 $0 Per capita GDP (2011 PPP $) $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 $0 Labor Productivity Malawi Linear (Labor Productivity) Per capita GDP (2011 PPP $) Capital Intensity Malawi Linear (Capital Intensity) FIGURE 40: MALAWI'S LABOR PRODUCTIVITY AND CAPITAL INTENSITY FIGURE 41: TFP IS HIGHER IN MALAWI THAN COUNTRIES AT ARE HIGH BUT CAPITAL PRODUCTIVITY IS LOW SIMILAR LEVEL OF DEVELOPMENT36 TFP relative to Malawi (Malawi=100%) Value added divided by capital 400% 480% TFP (relative to Malawi) 350% 400% 300% 320% 250% 200% 240% 150% 160% 100% 80% 50% 0% 0% $0 $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 Per capita GDP (2011 PPP $) TFP Malawi Linear (TFP) Capital Productivity Malawi Linear (Capital Productivity) Source: Authors’ calculation based on data from WBES Note: All data points are for the median firm on each measure of performance. For presentational purposes the chart is shown only for countries with per capita GDP between $0 and $12,000. Countries with GDP per capita over this amount are, however, included when we calculate the linear projection 35 We do this because many, although not all of the measures, appear to vary consistently with per capita GDP 36 Note for Figure 40: For presentational purposes the chart is shown only for countries with per capita GDP between $0 and $12,000. Countries with GDP per capita over this amount are included when we calculate the linear projection. The country average estimates are from the LAD regressions, although the results are virtually identical when coefficients from the OLS and frontier regressions are used instead. For example, labor productivity, capital intensity, and per worker labor costs all appear to increase as per capita GDP increases. In contrast, other measures such as unit labor costs do not appear to increase consistently with per capita income (Clarke, 2012). 80 Graphs below present plot of unit labor costs, employment growth over a period of three years, and provision for formal training. Like in the analysis above, if a point is above the regression line, the partial productivity measure is greater than would be expected given the county’s per capita income. If it is below the regression line, the opposite is true. FIGURE 42: DESPITE HIGH LABOR COSTS, UNIT COST IN MALAWI FIGURE 43: OVER THE PAST THREE YEARS, EMPLOYMENT GROWTH REMAINS REASONABLE WAS SLOW AMONG MANUFACTURING FIRMS Employment growth (ave. annual grwoth, Unit labor costs (% of value-added) previous 3 years) $0 $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 $0 $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 Per capita GDP (2011 PPP $) Per capita GDP (2011 PPP $) Per worker labor costs Ave. growth rate Linear (Ave. growth rate) FIGURE 44: FEW FIRMS IN MALAWI PROVIDE WORKERS WITH FORMAL TRAINING NOTES: 90% 80% % of firms with formal training programs 70% 60% 50% 40% 30% 20% 10% 0% Per capita GDP (2011 PPP $) % of firms with training program Malawi Linear (% of firms with training program) Source: Authors’ calculation based on data from WBES 81 ANNEX 3: ESTIMATING TOTAL FACTOR PRODUCTIVITY USING ENTERPRISE SURVEY DATA This annex 3 describes how total factor productive (TFP) for the cross-country TFP comparisons are estimated and how the within-Malawi-comparisons for different types of firms are estimated. It also describes how changes to the investment climate that might improve TFP in the country are estimated. A. Methodology to Estimate Total Factor Productivity Total factor productivity (TFP), or Technical Efficiency (TE), is the residual from a regression of the log of output (either value-added or revenue) on labor, capital, and other intermediate inputs. Following Caves and Barton (1990), this annex uses value-added rather than sales as the dependent variable. The estimation assumes a Cobb-Douglas Production function:37 Y  A KL (2) i i i i Y is value-added for firm i, K is a measure of capital (e.g., the book value or replacement value of capital), L is the number of workers and A is TFP or TE. Because constant returns to scale are not imposed, productivity can differ by firm size. When A is larger, the firm produces more output with the same amount of capital and labor. Taking natural logs of both sides implies: ln Yi      ln  Ki    ln  Li   i (3) Where: Ai  evi  e  i That is, firm i’s productivity is equal to a constant, μ, and a firm-specific measure of productivity, ε i. It is easy to generalize this into a more general ‘augmented’ production function where the error term is : vi    FCi   ICi  i (4) FCiis a vector of variables representing the characteristics of firm i and IC i represents the investment climate that firm i faces. This implies that: ln Yi      ln  Ki    ln  Li   FCi   ICi   i (5) To allow productivity to be different in different countries, the regressions include a vector of country-year dummies ( c ).The country-year dummies control for country-level differences that might affect productivity.38 For example, firms in some countries might be more productive if infrastructure is better quality, the rule of law is stronger, workers are more highly skilled or better educated, or the country is more economically or politically stable than in other countries. Although some of these differences in parts of the empirical analysis are controlled when investment climate variables in the regressions are included, it is difficult to control for all aspects of the investment climate. By comparing the coefficients on the country dummies, average levels TFP in Malawi with TFP in other countries39 can be compared. This implies: 37 Itis possible to make other assumptions about the functional form of the production function (e.g., to assume a trans-log production function), although this does not appear to have a significant impact on the results in most cases. See, for example, Clarke (2014). 38 That is, for countries with multiple surveys we include separate dummies for each survey. This allows macroeconomic factors that might affect TFP to differ over time. For brevity, we will usually refer to these as ‘country dummies’ rather than ‘country- year’ dummies unless we want to explicitly remind the reader that they are country-year dummies. 39 For the LAD regressions, the coefficients are better thought of as median levels of productivity than average levels. 82 ln Yi   c   ln  Ki    ln  Li    FCi   ICi  i (6) Equation (5) can be estimated by Ordinary Least Squares (OLS) as long as the basic assumptions of the linear model are met. One important assumption is that capital and labor must be uncorrelated with the error term. This implies any shock that affects productivity must be uncorrelated with the firms’ capital and labor choices. This would be violated if, for example, managers were aware of something that affected productivity and decided to hire or fire workers or invest in new machinery and equipment to take advantage of a temporary or permanent increase in productivity. Managers might do this, for example, if they received technical advice from one of their suppliers or buyers that improved their firm’s productivity and then the manager decided to hire more workers to take advantage of this improved know-how. Characteristics of the firm and the investment climate can also be directly included in the OLS regression as long as these characteristics are exogenous.40If becoming an exporter makes a firm more productive (e.g., through exposure to foreign markets or greater competition) then a dummy variable indicating that the firm is an exporter could be included in the regression if productivity did not affect the firm’s decision to become an exporter. This would be violated, however, if, for example, a firm became more productive and decided that this meant that it could start exporting.41 Similarly, an investment climate variable could be included representing how much the firm pays in bribes if productivity did not affect how much the firm pays. If, for example, corrupt bureaucrats target firms that are especially productive—perhaps because productive firms are more profitable—then this assumption would be violated.42 Rather than including firm or investment climate characteristics directly in the model, it is possible to first estimate equation (2) through OLS or another more robust estimation method, obtain estimates of TFP by calculating ε for each firm from equation (2) and then regress the residuals on the firm and investment climate characteristics (e.g., estimating equation (3)). An advantage of this approach is that if panel data is available it might be possible to estimate equation (2) using a robust technique such as the method suggested by Levinsohn and Petrin (2003) and then use something such as 2SLS in the second stage if one of the firm or investment climate characteristics were thought to be endogenous.43 The drawback of this second approach is that if the firm level or investment climate characteristics are correlated with the amount of labor and capital that the firm uses (i.e., if the manager’s knowledge about the investment climate affects the firms use of labor or capital) then estimates of the coefficients in equation (2) will be biased.44 As a result, TFP (i.e., the error terms) will be estimated incorrectly and the coefficients from the second stage will also be biased. It seems this will often be the case. Escribano and Guasch (2005), argue that “this is almost always the case since the inputs are correlated with the Investment Climate (IC) variables and least squares estimators of [equation 2] are inconsistent and biased. ”For this reason, estimation is done in a single step in this report. Another concern about OLS is that outliers can have a significant effect on OLS estimates. If some firms misreport their sales, capital, or workers, this can affect estimation considerable. Given that many of the firms in the analysis do not report data directly from company accounts —and that many small firms in developing countries might not even keep detailed accounts —this can be a serious problem. In the 40 This also assumes that there are no omitted variables correlated with both the characteristics of the firm or investment climate and productivity. 41 There is a large literature on whether exporting improves performance (learning-by-exporting hypothesis) or whether only productive firms can export (self-selectivity hypothesis). The large literature on this topic is summarized in Tybout (2003) and Bernard and others (2007). 42 Svensson (2003) find more profitable firms pay more in bribes than other firms do. 43 Gatti and Love (2008) do this allowing access to credit to be endogenous in the second step. 44 This is due to omitted variable bias. It is discussed in more detail in Chapter 7 in Kumbhakar and Lovell (2000) and in Escribano and Guasch (2005). 83 Enterprise Survey for Malawi, for example, only about 50 percent of firms report that they have audited accounts and only about 29 percent of firms always reported numbers directly from their accounts. 45 Outliers can be dealt with in several ways. One way is the estimate the equation with a robust estimation method such as a Least Absolute Deviations (LAD) estimator, which weights outliers less heavily than OLS does.46 Another approach is to drop outliers. In OLS and frontier estimation, which are less robust to outliers, we exclude outliers that report value-added per worker or capital per worker more than three standard deviations from the mean for that country. In most countries, this drops about 10 percent of the sample. In Malawi, about 10 percent of firms are dropped.47 As discussed below, however, the main results are similar in the LAD regressions, which include outliers, and the other regressions, which exclude them. In addition to concerns about outliers, other concerns have been discussed in the literature. These include the functional form of the error term (ε) and whether the error term is correlated with capital or labor. There are several methods that have been proposed regarding the functional form of the error term. Stochastic frontier analysis allows for two error terms, a one sided term assumed to have a half normal distribution, ν, representing technical efficiency and a two-sided normally-distributed error term, , representing temporary shocks to productivity and measurement error. The frontier model is estimated using maximum likelihood estimation. In the analysis, we use the LAD estimators as our base estimator. We also estimate the model using standard OLS estimators and stochastic frontier analysis as robustness check. In practice, the results are similar in the two models. As discussed below, the coefficients on the investment climate and firm characteristic variables are similar in terms of size and statistical significance in the three models. Further, the estimates of average TFP in different countries are also not sensitive to the way that we estimate the production function. The correlations between the country-level estimates of productivity from the three methods are between 0.966 and 0.981.48 That is, countries that appear more productive using one estimation method also appear more productive using other methods. In the text, we therefore focus on results using the LAD estimator. The TFP measures are also very highly correlated with labor productive with correlations between 0.913 and 0.952, indicating that sectoral differences and differences in capital intensity explain relatively modest differences in labor productivity. TABLE 24: CORRELATION BETWEEN COUNTRY LEVEL ESTIMATES OF TECHNICAL EFFICIENCY Median OLS Frontier Labor Productivity Median 1.000 0.981 0.966 0.926 OLS 1.000 0.967 0.952 Frontier 1.000 0.913 Labor Productivity 1.000 Source: Author’s calculations based on data from the WBES. 45 An additional 34 percent reported some numbers from their accounts. 46 Due to concerns about outliers, LAD estimators are often used when estimating production functions. See, for example, Greene (2002, pp. 449-450). 47 This does not include firms that refused to report productivity data (i.e., it is the percent of firms that reported enough data to calculate total factor productivity that were more than 3 standard deviations away on those two dimensions for firms that reported data). 48 Using Enterprise Survey data from a slightly smaller group of countries, Clarke (2014) shows that the coefficients on the country dummies are very similar under a broad range of different assumptions. In addition to using OLS, LAD, and frontier estimation, Clarke (2014) also estimates translog production functions, includes additional sector dummies and interaction terms, and estimates a restricted model. 84 Note: Based on calculations from regressions with country dummies, sector dummies, and sector specific factor intensities but no investment climate or firm characteristics included. A broader problem is that things that affect productivity might affect firm managers’ choices regarding capital and labor. If this is the case, OLS, stochastic frontier estimation and LAD estimation will all produce biased estimates of the coefficients. Although several methods have been proposed to deal with this, they require panel data (Levinsohn and Petrin, 2003; Olley and Pakes, 1996) —something that is not available for most Enterprise Surveys. A final concern is the analysis includes firms from many different areas of manufacturing. The production function in equation 7 assumes all firms in all different areas use the same production technologies. A more flexible estimation technique that allows firms in different sector to use different production technologies might be preferable to a technique that assumes all firms use the same production technologies. This can be done mechanically by including sector dummies in the regression and interacting the dummies with labor and capital to allow different labor and capital intensities across subsectors of manufacturing. The augmented production function then becomes:   ln  Yij   c    j   j log  Kij    j log  Lij    FCij   ij  ij j (7) The coefficients on labor and capital, β and γ, are assumed to be different in different subsectors. Subsector dummies, α, are also included to allow for systematic differences in productivity.49 Models that include the additional dummies and the interaction terms are sometimes referred to as ‘unrestricted models’ while the models that assume identical production technologies are referred to as ‘restricted models.’ In practice, including the extra terms does not appear to have a significant effect on the estimates of average TFP across countries (Clarke, 2014). B. General Methodological Issues There are some well-known problems with this methodology. These include the following: 1. To make cross-country comparisons of TFP, value-added and capital have to be denominated in a common currency (e.g., US dollars in these examples). Because these two variables are denominated in local currency in the survey, cross-country comparisons of TFP are vulnerable to exchange rate fluctuations. If the exchange rate is overvalued relative to its long-run equilibrium then TFP might look artificially high. Although this can make it difficult to interpret differences in TFP between countries, this does not affect coefficients on the investment climate or firm-level variables. That is, because the estimation uses natural logs of value-added and capital, the country- year dummies will control for exchange rate fluctuations. 2. The model assumes firms in different countries in the same sector use similar technologies. That is, it assumes that garment firms in Bangladesh have the same production function as garment firms in Slovenia. It also assumes that firms within broad sectors use similar technologies (e.g., that firms that make t-shirts have similar production functions to firms that make dresses). In practice, the results for the country averages appear robust to including additional sector dummies and interaction terms.50 49 The analysis in this paper includes 12 sub−sector dummies and the related interaction terms. The sub−sectors are garments; textiles; food and beverages; chemicals; metals and machinery; electronics; non−metal minerals; wood and furniture; paper and publishing; plastics; automobiles and parts; and other manufacturing. As noted below, the results are not highly sensitive to using more detailed breakdowns. 50 Clarke (2014), in particular, includes 176 subsector dummies, measured at the 4−figure ISIC classification level, and interacts the dummies with labor and capital (i.e., allowing different capital and labor intensities in each subsector). The results are similar to when 12 subsector dummies are included. 85 3. Capital is more difficult to measure than labor for both theoretical and practical reasons. In addition to the standard problem of dealing with depreciation, few firms keep audited accounts and few managers report figures directly from their accounts. Most firms will have purchased their machinery and equipment over a long period —and sometimes in the distant past—meaning that managers find it harder to estimate the value of their machinery and equipment than they do to estimate things such as sales or expenditures over the past year. Because of this, capital is probably measured less accurately than value-added or the number of workers the firm has. Even if the mismeasurement is random, the coefficients will be biased. Because the coefficients are biased, TE, which is the residual from the regression, will also be mismeasured when capital is mismeasured. 4. Ideally the measure of output would be a physical measure of output. In practice, because it is difficult to obtain physical measures of output, most TFP analyses, including those using Enterprise Survey data, use sales (i.e., output multiplied by unit price) as the dependent variable. Because production is affected by price as well as quantity, these functions are sometimes referred to as sales generating functions rather than production functions. 51 With firms producing heterogeneous products, this can be a problem if some have market power. That is, a firm with market power that charges high prices (e.g., monopolists) would appear more productive than a similar firm in a competitive market that charges lower prices.52 5. Because estimates are calculated in a regression framework, it is less straightforward to calculate TFP than labor productivity. One issue is that estimates of TFP for groups of firms do not have natural units. For cross country comparisons, TFP is shown as percent of TFP in Malawi. For other groups (e.g., exporters) differences are presented in terms of a base category (e.g., non-exporters). 6. As noted by Escribano and Guasch (2005), there is no single accepted approach to estimating TFP. For this reason, following Escribano and Guasch (2005), we estimated the model in several different ways to check the robustness of results. We therefore estimate the model in various ways making different assumptions about the error terms (i.e., we present estimates from stochastic frontier, OLS, and LAD models. 7. Recent studies have noted that inputs in the production function (labor and capital) are endogenous (Levinsohn and Petrin, 2003; Olley and Pakes, 1996). This can affect the estimation of TFP. With panel data, it is possible to control for this using sophisticated econometric techniques, instrumenting for inputs with intermediate inputs or investment. In this case, without a long panel, these methodologies cannot be implemented. C. Methodology Issues Related to Investment Climate Variables As noted earlier, investment climate variable can be directly included in the productivity regressions as long as they are exogenous. That is, they can be included as long as productivity does not affect the measure 51 See,for example, the discussion in Pakes (2008). 52 See,for example, the discussion by Levinsohn (2008) on the Escribano-Guasch methodology (Escribano and Guasch, 2005; Escribano and others, 2008; Escribano and others, 2005). 86 of the investment climate (exogeneity) and as long as there are no omitted variables that are correlated with both productivity and the investment climate ( no omitted variables). To avoid the omitted variable problems, the regressions include many investment climate variables. These include variables to control for the burden of regulation, petty corruption, the cost of crime and security, the reliability of electricity supply, the availability of bank credit, the ease of Internet access, access to foreign technologies, and the quality of workers. The regressions also include country dummies, which should help control for other omitted investment climate variables that might affect firm performance. All investment climate variables included in the analysis are objective. Although the Enterprise Surveys also include subjective questions about managers’ perceptions about the investment climate, these variables are not included in the regressions because these variables might not be comparable across firms, regions, or countries. Various biases have been shown to affect subjective questions in ways that might affect the reliability of regressions coefficients. For example, Kaplan and Pathania (2010) show that macroeconomic performance affects managers’ perceptions about unrelated aspects of the investment climate. Similarly, Clarke (2010) shows that a power crisis in South Africa appears to have affected managers’ perceptions about unrelated areas of the investment climate. If macroeconomic stability or other things that affect perceptions also affect firm performance, this could bias coefficients on subjective variables. In addition to concerns about omitted variables, it is also possible that some investment climate variables are endogenous. For example, better performing firms might be more able to pay for training programs or internet access and find it easier to get bank loans. If these variables are included in productivity regressions, their coefficients—and coefficients on others—would be biased. Even if productivity does not affect the investment climate variables directly, omitted firm-level variables such as the manager’s ability might be correlated with firm productivity and with aspects of the investment climate such as how well the firm handles power outages or corruption. Because of these concerns, it is difficult to include investment climate variables directly in firm-level productivity regressions. Because of concerns about endogeneity, we do not include the investment climate variables directly in the productivity regressions. Instead, we average the investment climate variables across similar firms in the same city and include the averages in the regression. The averages are taken over all firms of similar size and in the same sector as the firm in each city. These city-sector-size averages are called as ‘local averages’ in the analysis. That is, we use other firms’ experiences in each area of the investment climate as proxies for the firm’s experiences. If similar firms in the same location tend to have similar experiences, this might be a reasonable proxy. In addition, because a single firm’s responses are unlikely to have a large impact on the average responses of all similar firms in the city, the local average will be less affected by that firm’s performance than is the firm’s own response. We include the averages directly in the equation rather than using them as instruments. There are three reasons for this. First, in many cases, the local average might be a more appropriate measure of the investment climate than the firm’s own response. For example, firms that refuse to pay bribes will be affected by corruption if it means they fail to get needed licenses or lose government contracts. That is, the overall level of corruption in a region will affect the firm’s performance whether or not the firm pays bribes. Second, because the regressions include many investment climate variables, 2SLS might be problematic due to weak instruments and other problems. Finally, using 2SLS would greatly reduce sample size. Firms often refuse or fail to answer the investment climate variables. When 2SLS is used, any firm that failed to answer any question needs to be dropped. In contrast, local averages can be calculated for almost all firms in the sample.53 Given that there are so many investment climate variables in the regression, many observations would end up being dropped. his approach, replacing the firm’s own values with the local 53 To avoid sample loss, when no firm in the city-size-sector group answers the question, we use the local average for all firms in the same city and sector. Similarly, if no firms in the city-sector group answer the question, we use the city average. 87 averages, is standard in the literature—many studies that have looked at how the investment climate affects firm performance have taken this approach.54 A final issue related to the investment climate variables relates to a final group of questions where managers were asked to estimate losses due to investment climate problems: losses due to crime and security costs; bribe payments; and losses due to power outages. The questions that these three variables are based on questions that allow the managers to answer either as a percent of sales or in monetary terms. Although this should not matter in principles, earlier studies have shown that managers who answer in monetary terms report significantly lower losses than managers who answer as a percent of sales (Clarke, 2012; Malomo, 2013). Moreover, these differences do not appear to be due to observable or unobservable differences between firms whose managers answer as a percent of sales and firms whose managers answer in monetary terms (Clarke, 2012). Although it is not clear whether managers who report as a percent of sales over-report losses or managers who report in monetary terms under-report, it is important to ensure that managers’ responses are comparable. To do this, we regress the reported losses on a complete set of country dummies and a dummy indicating the manager answered in monetary terms. We then calculate responses assuming the manager answered in monetary terms. D. Total Factor Productivity Related to Other Countries As a first exercise, we estimate equation 8 without any investment climate variable or firm characteristics. These regressions, however, include country dummies that we can use to compare average levels of productivity across countries. To calculate average TFP, we look at the coefficients on the country dummies. To calculate average TFP in each country relative to TFP in Malawi in 2014, we calculate the following equation: TFP  exp  c  (8) Where c is the country dummy for country c. This formula was proposed by Halvorsen and Palmquist (1980). For the OLS and frontier regressions, this can be interpreted as the average TFP in the country after controlling for sector and capital and labor intensities. For the LAD regressions, it can be interpreted as median TFP in the country. Since the omitted country dummy is for Malawi, the estimates give average or median TFP relative to average or median TFP in Malawi. TFP is lower in Malawi than in most of the comparator countries. Although TFP is slightly higher for the median firm in Malawi than in Mozambique or Tanzania (57 and 77 percent of TFP in Malawi), it is lower than in the other regional comparator countries. Moreover, although labor productivity in Malawi was higher than in Zambia, TFP is lower. This suggests that the higher labor productivity is due to firms in Malawi being more capital intensive than firms in Zambia rather than more productive. Although total factor productivity appears low in Malawi, this probably partly reflects Malawi’s low per capita income.55 As noted earlier, per capita income is lower in Malawi than in the comparator countries. Because total factor productivity tends to be higher in countries where per capita income is greater, it might not be surprising that TFP is relatively low in Malawi. To see whether this is the case, we compare TFP in Malawi with TFP in all countries where Enterprise Surveys have been completed. We then plot TFP against per capita income. The upward sloping regression 54 Dollar and others (2005), Aterido and others (2009), Harrison and others (2014), Fisman and Svensson (2007), and Clarke and others (2015) take this approach in their analyses. Xu (2011) discusses this approach in more depth. 55 Malawi’s per capita GNI was only $815 in purchasing power parity (PPP) US$ in 2014 (World Bank, 2015). This is lower than in all but three of the 188 countries where the World Bank has computed per capita GNI. 88 line confirms that TFP is higher in countries where per capita income is higher (see Figure 40 in Annex 2). TFP is higher in countries above the regression line than would be expected given the country’s per capita income. 89 ANNEX 4: TFP AND INVESTMENT CLIMATE Table below shows results from the base quantile regression model in Clarke (2015), which analyzed the effects of investment climate variables on TFP.56 The coefficients on the investment climate variables are statistically significant and have the expected signs. Firms are more productive in areas where they spend less time dealing with regulation, have smaller losses due to power outages and crime, pay less bribes, and have better access to credit. Further, the coefficients on the percent of firms with their own websites, the percent of firms with formal training programs, and the percent of firms that license foreign technologies are positive and significant. These results suggest that productivity is also higher in areas where firms have easier access to foreign technology, where internet use is higher, and where workers have more human capital. These results are mostly robust in the two alternate models, namely the base quantile regression model and ordinary least squares (OLS).The two main exceptions are that the coefficient on time dealing with regulation becomes statistically insignificant in the OLS model. It remains about the same size, however, as in the quantile regression. The other main difference is that the coefficient on bribes becomes statistically insignificant and smaller in size in the frontier regression. GENERAL COMMENTS ON THE TFP MEASURE Apart from the statistical considerations above, some authors have highlighted some potential drawbacks of the TFP measure. For example, Felipe (2008) identified the following: • The measure is calculated on the assumptions of profit maximization, competitive markets and constant returns to scale, which are crude and are not always considered by researchers in calculation of the measure. • There are challenges with apportioning growth in output to respective factors in growth accounting exercises due to the fact that growth is the result of the complex interaction of a number of different factors. • There are interpretation challenges whereby some researchers regressions of TFP as a dependent variable on many different variables just to identify some statistically significant variables that could explain the composition of the TFP measure. • Most TFP estimates are biased downward since the actual factor shares from national accounts are seen to have been affected by technical progress, rendering the TFP estimates inaccurate. • Savings rates and population growth do not clearly affect the steady-state growth rates of per capita output and hence productivity and long-term growth as posited by Solow’s neo -classical growth model, which depends on a number of very restrictive assumptions. • The aggregation of factors in the context of aggregate production functions is problematic as it relies on very stringent conditions which are probably not attained in practice. • In spite of these suggested challenges, TFP has been an important measure for explaining increases in productivity and hence long term growth over the years. As put by van Ark (2014), TFP does not suffer from diminishing returns as homogenous inputs typically do, and it also represents spillovers or externalities that arise from returns on inputs that go beyond what the investor can internalize. In other words, they relate more to investors’ operating environment, which is of concern to this paper. Under strict neo-classical assumptions, growth in TFP measures technological change. In practice, it also measures the impact of other inputs that are not measured such as research and development (R &D), human capital skills development, imperfect competition, increasing returns to scale and other externalities related to investment, the reallocation of market shares across firms and other measurement errors arising from measuring outputs and inputs in the productivity process. In a nutshell, TFP is seen to be affected by factors such as: 56 is based on the base regression results in Clarke (2015). Clarke (2015) discusses the robustness of the results and presents results from alternative models. 90 • Intangible investments such as education and skills, R&D, patents, licenses, organizational change and product marketing; • Skills, motivation and competencies – in part affected by intangible investments above; • Innovation and technological change – in part also affected by the intangible investments above; and; • Markets, institutions and regulations. 91 TABLE 25: BREAKDOWN OF PARTIAL PRODUCTIVITY DATA BY FIRM TYPE (1) (2) (3) Value added (natural log) Quantile OLS Frontier regression regression Observations 31,297 28,053 28,053 Sector dummies Yes Yes Yes Country dummies Yes Yes Yes Sector-specific factor intensities Yes Yes Yes Local Averages Time dealing with regulation (%) -0.002* -0.002 -0.001** (-1.92) (-1.50) (-2.04) Bribes (% of sales) -0.014*** -0.009*** -0.006 (-2.67) (-10.38) (-1.40) Losses to crime (% of sales) -0.033*** -0.044*** -0.028*** (-3.83) (-3.01) (-3.75) Losses due to power outages (% of sales) -0.022** -0.029** -0.031*** (-2.43) (-2.46) (-3.94) % of firms with bank credit 0.112*** 0.129*** 0.129*** (3.84) (3.26) (5.00) % of firms with own website 0.308*** 0.326*** 0.313*** (10.28) (7.04) (11.90) % of firms that license foreign 0.145*** 0.146*** 0.136*** technologies (3.92) (3.22) (4.13) % of firms with training programs 0.127*** 0.082*** 0.086*** (4.46) (2.94) (3.45) Firm Characteristics Firm is an exporter (dummy) 0.186*** 0.174*** 0.169*** (10.93) (8.12) (11.40) Firm is foreign-owned (dummy) 0.281*** 0.173*** 0.185*** (10.95) (6.00) (7.85) Age of firm (nat. log, years) 0.015 0.026** 0.020** (1.63) (1.99) (2.52) Constant 6.401*** 6.258*** 7.473*** (41.77) (48.57) (54.39) Adjusted R-Squared --- 0.782 --- Source: Author’s calculations based upon data from the World Bank Enterprise Survey. ***, **, * indicate statistical significance at 1%, 5% and 10% levels Note: T-statistics in parentheses. All regressions include capital (natural log of sales value), number of workers (natural log), sector dummies, and country dummies. The sector dummies are interacted with the capital and labor variables to allow capital and labor intensities to vary across sectors (see equation). The local averages are average values of IC variables for similar firms (see description in text). Standard errors in the OLS regression are clustered at the country level. Clustering is not possible in the quantile or frontier regressions. Source: Authors’ calculation based on data from WBES 92 ANNEX 5: WHAT IS AN ENTERPRISE SURVEY An Enterprise Survey is a firm-level survey of a sample of private sector firms in an economy, following standard methodology and approach, to obtain firm level performance and constraints . The World Bank undertakes periodic enterprise surveys throughout developing countries, in order to provide policymakers with a better understanding of how different factors of the business environment are facilitating or constraining efficiency and productivity of firms and hence a country’s prospects for reaching its potential with respect to employment, production and welfare. Since 2002, the World Bank has collected this data from face-to-face interviews with top managers and business owners in over 155,000 companies in 148 economies/countries. Survey data is collected by private contractors on behalf of the World Bank57, and covers business environment topics including Access to Finance, corruption, infrastructure, crime, competition, and performance measures. World Bank enterprise surveys follow a standard format and methodology. The survey covers firm characteristics, gender participation, Access to Finance, annual sales, costs of inputs/labor, workforce composition, bribery, licensing, infrastructure, trade, crime, competition, capacity utilization, land and permits, taxation, informality, business-government relations, innovation and technology, and performance measures. The Enterprise Surveys adopt a stratified random sampling technique. The strata include firm size, business sector, and geographic region within a country. Firm size levels are 5-19 (small), 20-99 (medium), and 100+ employees (large-sized firms). Panel data are usually obtained, by interviewing same firms across multiple years. This has been a priority in recent Enterprise Surveys. 57 Due to sensitive survey questions addressing business-government relations and bribery-related topics, private contractors, rather than any government agency or an organization/institution associated with government, are hired by the World Bank to collect the data 93 ANNEX 6: ENTERPRISE SURVEY IN COMPARISON TO DOING BUSINESS METHODOLOGIES Enterprise Surveys (ES) and Doing Business (DB) are used to assess the quality of business environment. An ES is a firm-level survey of a representative sample of the private sector in an economy, covering a range of business environment topics such as access to finance, corruption, infrastructure, crime, competitions and performance measures. DB focus on measuring the complexity of business of business regulations and quantifying the ease of doing business across countries through indicator sets and rankings. The indicators use common transactions such as starting a business or studies. Table below provides some key features of the two methods. Global coverage Covers 139 economies Collects data annually for 190 economies Data Source Firm level data. Unit of observation is Collects information through surveys and enterprise. Business owner or top administered to local experts on the subject manager is usually given face to face matter or business transaction such as interview. Business surveyed include lawyers, accountants, and architects. The manufacturing, retail, construction, information is confirmed through the transport, communication and other underlying laws and regulations services Geographical The sample design aims to include the Collects data only for the main (most coverage within an main cities/regions of economic activity populous business city) economy Information Objective data on the business Measures 11 business regulation topics: gathered environment as experienced by firm, starting business, dealing with construction performance measures, firm permits, getting electricity, registering characteristics, and perceptions property, getting credit, protecting minority regarding obstacles to growth investors, labor market regulations, paying taxes, trading across borders, enforcing contracts, and resolving insolvency. Data Stratified Random sampling Standardized case-studies that relate to a common business situation Measurement Measures what happens to existing Measures what a standardized firm should firms, such as the actual experiences of a expect if everything was done according to firm regarding the payment of taxes, the official legal requirements and costs in number of meetings with tax officials place. and the incidence of bribery. The survey also asks how much of an obstacle a challenge is to business growth. Assumptions Measures what happens in practice in the Assumes firms are aware of and comply normal course of business. with all formal regulations, waste no time in collecting information and all regulations are enforced. Usage Can be useful to identify potential areas Can be used to identify areas for reform of reform in the business environment as based on bottlenecks or weaknesses in well as ass the impact of reform on specific areas of private sector regulation, businesses gain insights and learn from practices of other economies around the world. Source Enterprise Survey (World Bank, 2017c) 94