1

PRIVATE INVESTMENT AND LABOUR DEMAND IN SUB-SAHARAN AFRICA

AGYEI, SAMUEL KWAKU

PhD FINANCE STUDENT (10292234)

2ND DRAFT 3/04/14

2

1.0 Introduction

The development of every nation is the ultimate goal of all economic, social and political policies. Job creation contributes to development by boosting living standards, raising productivity, and fostering social cohesion (World Bank, 2013). Unfortunately, however, some 200 million people (predominantly young people) are unemployed. The estimates are even gloomier because apart from the fact that about 600 million new jobs would have to be created by 2020 (mainly in Africa and Asia), these jobs need to be good (International Financial Corporation-IFC, 2013). In spite of this need, the global credit crunch has worsened an already deteriorating global employment condition. Its effect on employment has been a dipping global employment level, further aggravating an age-long problem. Nickell (2010) asserts that the worldwide credit crunch and collapse of aggregate demand should be blamed for the recent rise in unemployment. Earlier, Kessing (2003) argued that the effect of the economic crises on public sector employment has been a reduction in real wage and not the level of employment. But recently, The International Labour Organisation – ILO (2013) report indicated that global employment trends do not only show a rise in unemployment but with significant regional differences. The report further states that five years after the outbreak of the global financial crisis, labour markets remain deeply depressed and unemployment has started to rise again as the economic outlook worsens. The African economy was not insulated from the negative effect of the crunch including that of the growing global unemployment challenge, largely through spillovers (ILO, 2013). In Africa, even in a regime of increasing economic performance, Emery (2003) warned of a decreasing employment content of growth and increasing inequality over the preceding few decades. Thus employment generation still remains a global challenge especially for Africa which need not just create more jobs but good ones.

These have resuscitated a new search for fighting unemployment and its related problems. In the search for solutions for this global challenge, Guy Ryder - ILO director, advices (in ILO Report 2013) that “The global nature of the crisis means countries cannot resolve its impact individually and with domestic measures only.” He explained that the cloud of uncertainty surrounding investment and job creation means that countries need to take concerted actions to help resolve this growing global challenge. Investment is one of the traditional ways of curbing unemployment basically because manpower compliment or serve as a substitute for physical capital. Cherian (1996) argued that investment may be considered the most important component of GDP because (1) Plant and Equipment have a long-term effect on the economy’s productive capacity (2) Changes in investment spending directly affect levels of employment and worker’s incomes in durable goods industries and (3) supply and demand are sensitive to changes in investment, which is the most volatile component of GNP. In an analysis of the relationship between governance, transparency and private investment in Africa, Emery (2003) observed that private investment in a country had positive effects not only on growth but also on the incidence of poverty. Impliedly, private sector investment, including domestic and foreign direct private investment, when operated in a conducive environment, can be a key driver of economic development, job creation and inclusive growth. Consequently, the role of the private sector in solving this global challenge cannot be underestimated, especially in developing economies where about 90% of jobs are provided by the private sector (IFC, 2013).

Even though the Sub Saharan African region’s Unemployment rate, as at 2011, (about 8.8%) was better than that of North Africa (about 10.9%), Middle East (about 10.5%), Central and

3

South-Eastern Europe (about 9%), the performance of the region was about 2.4 percentage points worse than the global average. Also, most of the jobs in the SSA region seem not to be good, as the region was the second worse region in the world in terms of share of working poor. About 65% of total employment in 2011 was found to belong to this category. This situation is particularly worrying because it is more than double the global average (about 29%) (International Labour Organisation-ILO, 2012). Analyses of the changes in employment in the SSA region, over the study period, show some interesting results. Generally, the second decade of the study period (2000-2009) shows an increase in employment to population ratio from 63.77% (1990 – 1999) to 64.46%. Interestingly, while more females are joining the working populations (55.31% to 57.18%), the opposite can be said of their male counterparts (fell from 72.60% to 71.95%), when the two decades are compared. Apart from the fact that the total percentage of youth working fell (from 47.48% to 46.89%), the changes in female youth employment (increased from 42.93% to 43.10%) and that of male youth employment (decreased from 52.07% to 50.68%) is reminiscent of movements in total female employment and total male employment, when the first and second decades of the study periods are compared.

Meanwhile investment has seen some considerable improvement. Total investment in the second decade of the study period (2000 – 2009) showed a marginal increase from 19.72% (1990 – 1999) to 20.06% of GDP. There is also evidence of a gradual shift from government led investment to private sector controlled investment in the SSA. Public sector investment fell from 7.72% (1990 – 1999) to 7.10% (2000 – 2009) while private investment increased from 12.40% of GDP to 13.10% of GDP. All throughout the study period, private investment has accounted for the greater proportion of total investment (Appendix 1). Also, between 2001 and 2010 net flows of foreign direct investment in Sub-Saharan Africa totaled about US$33 billion—almost five times the US$7 billion total between 1990 and 1999—and export growth was robust (World Bank 2011). These figures reinforce the need for Sub – Saharan Africa to put in measures to get the best out of private investment.

In spite of this, little is known empirically, on the employment benefits of private investment. Discussions on the continent on private investment have largely been concentrated on how to attract private investment (Oshikoya, 1994; Mlambo and Oshikoya, 1999). In 2002, Devarajan, Easterly and Pack opened the argument box on whether it is the size of private investment on the continent that should be of grave concern or the productivity of private investment. They concluded that investment on the continent was not low because what even existed, at the time of their study, was largely not productive. Also, quite recently, AfDB,OECD,UNECA (2012) reported that even though FDI remains the largest external financial flow to Africa, the increase in investment in recent decades did not produce more inclusive growth or sufficient jobs as most of the finance went onto the hunt for resources. These studies seem to cast doubt on the actual benefits of private investment to the African continent. But to Kaplinsky and Morris (2009), SSA’s should use their resource endowment to get the maximum benefit from their investment relations with large state-owned Chinese firms and other large firms who seek to benefit from their resource endowment. These benefits, the study believes should include employment generation.

Dinh et al (2012) maintain that investment on the continent is low—less than 15 percent of gross domestic product, compared with 25 percent in Asia—and more than 80 percent of workers are stranded in low productivity jobs. They explain that in spite of this, the continent’s largest geographical bloc’s, Sub-Saharan Africa (SSA), economic performance is at a turning point after almost 45 years of stagnation. Between 2001 and 2010 the region’s gross domestic product grew at an average of 5.2 percent a year and per capita income grew

4

at 2 percent a year, up from –0.4 percent in the previous 10 years (World Bank 2011). International Monetary Fund (2013) adds that even with the exclusion of Nigeria and South Africa, most countries in Sub-Saharan Africa recorded increases in GDP. Unfortunately, however, even in periods of economic growth, employment generation is not a natural consequence unless conscious effort is made to make that growth beneficial to job creation (Inter-Agency Working Group – IAWG, 2012 and Heinsz, 2000). On employment in Africa, Asiedu (2004) looked at the determinants of employment in SSAs using data from foreign affiliates of US multinational enterprises in Africa, Sackey (2007) considered employment impact of private investment using a sample of SMEs from some African economies, Asiedu and Gyimah - Brempong (2008) studied the effect of liberalization of investment policies on investment and employment of multinational corporations in Africa and Aterido and Hallward-Driemeier (2010) used firm-level survey data from 104 developing economies which included 31 sub-saharan countries to find out whether investment climate fosters employment growth.

This study differs from that of Aseidu (2004), Sackey (2007), Asiedu and Gyimah-Brempong (2008) and Aterido and Hallward-Driemeier (2010) because it uses national data to assess the relationship between private investment (Not only from USA, foreigners or SMEs) and employment in SSA using the neoclassical labour demand model. The neoclassical labour demand theory predicts a negative association between labour cost, real factor cost and labour demand and a positive relationship between output and labour demand (Symons, 1982 and; Andrews and Nickell, 1982 and Sparrow, Ortmann, Lyne and Darroch, 2008). In spite of this, other researchers argue that a positive association between wage cost and labour demand is possible, through the aggregate demand channel, especially in recession (Keynes, 1936 and Michaillat and Saez, 2013). Thus, the study contributes to the discussion on the benefits of private investment to the African continent through the channel of employment generation and the neoclassical labour demand model. The study is concentrated on this particular area because: 1) even though job creation contributes to development, it has become a global challenge, especially after the crunch (World bank, 2013 and ILO, 2013); 2) it is a way of testing empirically for one of the pillars for assessing private investment impact (IAWG, 2012); 3) it is an appropriate channel to economic growth which has seen some improvements in Africa in recent times and seem to coincide with improvements in FDIs as well (World Bank, 2011; Dinh et al 2012); 4) employment seems to be an appropriate channel for ameliorating poverty (Emery, 2003) -one of the deep seated problems of the African Continent- and as a possible means of achieving millennium development goal – MDG 1; 5) it contributes towards the discussion on the effect of wage cost on labour demand and; 6) Insufficient labour demand is among the biggest causes of unemployment in Africa and indeed the biggest obstacles to youth employment on the continent (African Economic Outlook-AEO, 2012). In fact, the AEO report projects that a vigorous private sector is the most important vehicle for creating jobs for young people in Africa and that government must make it a priority to address the needs of the private sector.

2.0 Literature Review 2.1 Neoclassical Theory of Employment The study relates to the neoclassical theory of employment. This theory is popular in the area of demand for capital and labour (Van Reenen, and Bond, 2005). The classical employment analysis is based on the Market Law (Say, 1826) of exchange activity: "Supply creates its

5

own demand." Based on this view, classicists view unemployment as voluntary, temporary and partial. The theory explains that when labour supply is more than labour demand, employees are expected to accept pay cut so that employers can be motivated to employ more people in order to restore equilibrium in the labour market through the self- equilibrating tendency of the economic forces. The theory positions itself on the assumptions that the economy operates at full employment and that prices and wages are flexible. A firm's labour demand is then based on its marginal physical product of labour (MPPL).

The theory is criticized by Keynes (1936) on the grounds that the market law which forms the bedrock of the classical labour demand theory can only exist in a barter economy but not in a modern economy where money plays a major role as a medium of exchange. Thus, this casts doubt on the self-equilibrating tendency of the economic forces. Also Keynes argues that prices and wages flexibility does not always create equilibrium conditions. For instance a reduction in wage rates in periods of deep depression in an attempt to curb unemployment may worsen the unemployment situation because it would reduce aggregate demand. Thus, aggregate demand better explains employment than wage rate. Earlier researches failed to establish the predicted negative relationship between wage rate and employment as postulated by the neoclassical theory (Dunlop, 1938). Some of the reasons assigned to this were poor data quality (Symons and Layard, 1986), the size of expected output (Barro and Grossman, 1971) and costly adjustment of workforce (Sargent, 1978). But latter studies supported the theory albeit after accounting for the simultaneous effect of real price of raw materials and lags of real wage (Symons, 1982 and; Andrews and Nickell, 1982). These studies, therefore, placed particular importance on the effect of adjustment cost in determining the magnitude of wage shocks on labour demand (Kessing, 2003). It was Oi (1962) who set the tone for this area of research in labour demand. He explained that because of adjustment cost, labour is not a perfectly flexible factor of production. Kessing (2003) explains that changes in real wage rate do not sometimes have the desired impact on labour demand because changes in labour demand are affected by the adjustment process and its related cost. In other words, it takes time and money, for instance, to employ more people as a result of a fall in real wages because it requires training of new employees and expansion of existing facility. Weak labour mobility also stalls the adjustment process. Thus, firm response to wage changes is smaller in the short term than in the intermediate or long-term (Lichter , Peichl and Siegloch 2013). Job security provisions like high firing costs have the likelihood of improving long-run employment outlook for all workers. But then firing cost might reduce average labour demand for seasonal jobs. Turnover cost affects employment dynamics more than average employment (Lazear, 1990; Bentolila and Bertola 1990; Bertola, 1990 and; Bertola, 1991). Bertola (1991) concluded that firing cost may increase average employment while hiring cost reduces average employment. Hamermesh (1988) revealed that Employment levels only changes with large shocks and not with smaller shocks signalling that adjustment cost influences employment responses to policy changes. Goux, Maurin and Pauchet (2001) found in France that the firing cost of indefinite –term contracts is greater than their hiring cost and that because of the relative cost of hiring and firing workers, it is less costly to adjust the number of fixed term contracts than to adjust the number of indefinite-term contracts. From their study, the effect of hiring and firing cost is of particular importance to non-production workers than for production workers. Following from this, some studies have sort to estimate

6

the speed of the adjustment and generally concluded that the adjustment period is faster, within a year (Hamermesh, 1993). Exceptions to this include Nickell and Wadhwani (1991) Bentolila and Gilles St. Paul (1992) and Mairesse and Dormont (1985). The presence of adjustment cost makes it imperative to use dynamic models in modeling labour demand, in order to account for the inclusion of both contemporaneous and lagged values of the variables (Lichter, et al 2013). In spite of this, some studies have reported a negative relationship between real factor prices and labour demand. Symons and Layard (1984) concluded that real factor prices (real wage cost and raw material cost) are more important in determining the level of employment than aggregate demand. Also, Boug (1999) report, using data from Norwegian manufacturing and time series analysis that labour demand is influenced positively by production and negatively by the stock of real capital, real factor prices and total factor productivity. Pierluigi and Roma (2008) depict, based on data from the five largest Euro area countries, that job creation is enhanced through labour cost moderation even though the extent of enhancement varies across countries and sectors. When labour is assumed to be homogeneous, the cost of capital has no significant role in labour demand Köllő, Kőrösi and Surányi (2003). They also concluded that Production and labour costs are equally important explanatory variables of firm-level labour demand and that labour demand was more elastic downward than upward.

In South Africa, Sparrow, Ortmann, Lyne and Darroch (2008) reported that an increase in the cost of regular farm labour as a result of minimum wage legislation, resulted in a marked structural decline in the demand for regular farm labour. This notwithstanding, the debate on the exact impact of wage cost on employment seems to be far from being resolved. Very recently, Michaillat and Saez (2013) posit that when profits and wage are not equally distributed, a rise in wage rate may stimulate aggregate demand and reduce unemployment. This position casts doubt on the neoclassical theory but supports the Keynes view of employment. Knowledge about the wage elasticities of labour demand is important not only for economic research but also for policy analysis (Lichter et al, 2013 and Hameresh, 1993). Own wage elasticities of demand is not homogeneous across countries and that differences in institutional regulations play a major role in influencing this behaviour (Pierluigi and Roma, 2008 and; Lichter et al, 2013). In the short-run, the neoclassical model considers only wage cost as the main determinant of labour demand whiles wage cost, real interest cost and output are seen to influence long-run labour demand. This study contributes to the neoclassical theory by using disaggregated demand variable, private investment, to assess its impact on labour demand after controlling for other important factors like public investment and governance in Sub-Saharan Africa.

2.2 Empirical Literature Review This study tests, empirically, the potency of private investment in generating employment, as espoused in literature (Cherain, 1996 and Emery, 2003), using data from Sub-Saharan Africa. Research in this area have largely been concentrated on the employment impact of FDI (Driffield and Taylor, 2000; Henneberger and Ziegler, 2006; Karlsson, Lundin, Sjöholm, and He, 2007; Ndikumana and Sher, 2008; Mucuk and Demirsel, 2013; Habib and Sarwar, 2013) minimum wage (Neumark and Wascher, 2006; Neumark and Wascher, 2007; Herr, Kazandziska and Mahnkopf-Praprotnik (2009)); infrastructure investment and labour demand

7

(Garrett-Peltier, 2010; Pereira and Andraz, 2012). Wage elasticities of labour demand (Peichl and Siegloch, 2012). Technology, innovations and labour demand (Berman, Bound and Griliches (1994), Machin, Van Reenen and Ryan (1996), Van Reenen (1997), Falk and Koebel (2004), and Addison et al., (2008)); globalization and labour demand (Hijzen et al., (2005) and Hijzen and Swaim (2010)); ownership structure and labour demand (Barba Navaretti et al. (2003); capital structure and labour demand ( Funke, Maurer and Strulik (1999)) with very few concerntrating on private investment (foreign or domestic) in total on labour demand (Psaltopoulos, Skuras and Thomson, 2011).

Blomstrom, Fors and Lipsey (1997) revealed that in US larger foreign production is associated with lower parent employment, because of relatively low productive activities in parent country. They further explained that foreign production in developing economies and not developed countries was the main source of lower parent employment and that U.S. firms were enticed by lower wages in those regions. On the other hand, Swedish parents exhibited the opposite because their overseas production was more capital intensive than labour intensive. Impliedly, the study also brings to bear the employment effects of the nature of production systems. Capital intensive production systems have relatively low labour demand content and labour intensive production systems obviously have relatively high labour content. Thus the trade-off between the labour cost and physical capital cost is probably an important determinant of labour demand. Also, Harrison and McMillan (2004) postulate that increased capital mobility may be associated with negative labour outcomes for both the US and abroad. Garrett-Peltier (2010) assessed the employment impact of the US economy’s investment in renewable energy and energy efficiency and reported that this investment would lead to approximately three jobs being created in clean energy sectors for each job lost in the fossil fuel sector. From rural areas in southern Europe, Psaltopoulos, Skuras and Thomson (2011) show that private investment in agriculture showed a moderate impact on regional employment even though analysis of economy-wide jobs created showed that gross cost per job was significantly lower. Pereira and Andraz (2012) concluded in Portugal that investment in railway infrastructure does not only crowd in private investment and employment at the aggregate level but also show similar effects even at the regional level. Henneberger and Ziegler (2006) concluded that FDI can have both complimentary and substitutive effect on the labour market but the positive effect of fdi on employment is minimal. This partially supports Rosen’s (1969) and Griliches’ (1969) hypothesis that capital and skills are compliments. Masso, Varblane and Vahter (2007) investigated the employment effects of outward FDI on Estonia, a low-cost medium- income transition economy. They revealed that outward FDI had a positive impact on home country employment and the employment effects of domestic Estonian firms investing abroad was higher than that of foreign firms in Estonia investing abroad. To the researchers better economic management magnified these employment benefits and that the service industry performed better than the manufacturing firms. In Taiwan, overseas production is generally detrimental to domestic employment even though it has the tendency to increase domestic employment through increased domestic output, from enhanced competitiveness (Chen and Ku, 2003). Görg and Hanley (2005) stress that international outsourcing leads to significant decreases in plant level labour demand but outsourcing of services appear to have lesser negative impact than outsourcing of materials.

8

Karlsson, Lundin, Sjöholm, and He (2007) report a positive relationship between FDI and job creation and this is facilitated by some firm specific characteristics particularly access to export markets. This included direct employment by foreign owned firms and spillover effects on domestic firms. Thus, in the long run, FDI influences employment (Jayaraman and Singh, 2007)

Driffield and Taylor (2000) observe that increase in FDIs increase the demand for skilled labour directly and indirectly through technological spillovers which increases the relative skilled labour demand of domestic firms. Habib and Sarwar (2013) conclude, from time series analysis, that Foreign Direct Investment and per capita GDP have positive relationship with employment levels in Parkistan. Buzás and Klára (2006) assert that Foreign direct investment leads to job creation in Hungary. Malik, Chaudhry, and Javed (2011) opine that while fdi creates employment opportunities in Parkistan, trade openness and social and political dimensions of globalisation negatively affects employment. On the other hand, Although FDI affects development in general it may also lead to wage inequalities (ODI, 2002). Subsequently, OECD-ILO (2008) report indicates that workers engaged in MNEs tend to earn comparatively higher pay in their host countries. Mucuk and Demirsel (2013) investigated the relationship between FDI and unemployment in seven developing countries (Argentina, Chile, Colombia, Philippines, Thailand, Turkey and Uruguay) by using the panel data analysis. They revealed that FDI and unemployment have long-run relationship but their relationship is not homogeneous. While FDI was found to increase unemployment in Argentina and Turkey, it was found to reduce unemployment in Thailand. Klette and Førre (1998) argue that Research and Development investment and high-tech industries do not lead to job creation, using data from Norwegian Manufacturing firms. They cast doubt on the optimistic view about job creation in R&D intensive firms and high-tech industries (Katsoulacos, 1984). This partially supports Schumpeter’s (1943) creative destruction view. In Iran, government consumption and investment expenditures affect employment differently. While increase in government consumption expenditure is associated with decreases in production, employment and investment, increase in government investment expenditure - apart from industry and mining sectors - increases employment Fouladi (2010). In an attempt to meet the forecasted employment needs of some 100 million new jobs in the Middle East and North Africa (MENA), the World Bank (2004) reported that the new development model for that region should be based on a reinvigorated private sector. The report further explained that this model should also include better governance, greater integration into the world economy, and better management of the oil resources in the region. In the same Region, Ianchovichina, Estache, Foucart, Garsous and Yepes (2012) report that, for the following decade, if the region is able to meet the infrastructural investment needs of about 6.9 percent of gross domestic product, annual job creation (direct and indirect) would be about 4.5million.This is in spite of the fact that job creation from this channel alone would not be enough to solve the region’s unemployment. Also, in the 2000s and in the MENA region, Infrastructure investment in the construction sector was a major source of employment for the citizenry, as compared to other countries and sectors (World Bank,

9

2013b). The study further predicts that if the MENA region is able to commit to infrastructural investment estimated at $106million annually through 2020, it would generate approximately 2.5million infrastructure related jobs (Estache, Ianchovichina, Bacon, and Salamon (2013). This shows the researchers’ confidence in the employment potential of investment. In Africa, Ndikumana and Sher (2008) posit that the continent has witnessed an increase in FDI inflow but the effect of this resource inflow on economic development is yet to be ascertained. Even though the study recognised that one of the ways of assessing economic development impact of FDI is through its employment impact, the study did not do so. However, they concluded that FDI and domestic private investment are compliments in Africa rather than substitutes. Huang and Ren (2013) report from a survey of 16 Chinese enterprises in Johannesburg (South Africa) that these investments brought about job increment to the local people (local-skilled and unskilled labour) partially refuting international observers’ assumption that Chinese investment in Africa lacks significant employment content.

In Egypt, lack of access to the private sector is a recipe for unemployment. For instance, during the transition to private-sector-led economy in Egypt, unemployment was more prevalent among young educated women than their male counterparts, as a reflection of the fact that unemployment was becoming less generalized but more concentrated among groups that have a difficulty in accessing the private sector (Assaad, El-Hamidi, and Ahmed, 2000). Notably, the present study is particularly related to that of Asiedu (2004), Sackey (2007), Asiedu and Gyimah - Brempong (2008) and Aterido and Hallward-Driemeier (2010). Asiedu (2004) concluded that good infrastructure, higher income, openness to trade and an educated labour force have a significant positive impact on employment. Even though the study was on the determinants of employment in Sub-Saharan Africa, it only used data from foreign affiliates of US multinational enterprises in Africa. In addition, this study concentrates on the effect of private investment (foreign or domestic) on labour demand. It further analysis this effect from the point of view of total, male, female and youth employment and also controls for additional important variables such as governance and general consumption level. The current study also tests for the dynamic nature of labour demand because of adjustment cost effect. Sackey (2007) as part of a broader work of analysing the role played by small and medium scale enterprises (SMEs) in how private investment influences structural transformation and economic growth in Africa also considered employment impact of private investment using a sample of SMEs from some African economies. The study concluded, using a probit model, that investing firms are more likely to record net additions to employment than non-investing firms. Apart from the study not being domiciled solely in Sub-Saharan Africa, it did not also consider the employment impact of private investment beyond SMEs. Also certain key factors that we believe would influence the level of employment in the Sub-Saharan African region like governance were also left out. This study factors all these observations in SSA.

In another related study, Asiedu and Gyimah-Brempong (2008) studied the effect of liberalization of investment policies on investment and employment of multinational corporations in Africa. They revealed that while liberalization has a significantly positive effect on investment and its relationship on multinational employment is indirect. Data on employment was from employment of US affiliates in MNCs in Africa. Wage variable may be distorted because it carries with it the impact of the wage conditions that exist in the

10

mother company since it includes compensation of expatriates employed by the American affiliate in the host country. Aterido and Hallward-Driemeier (2010) used firm-level survey data from 104 developing economies which included 31 sub-saharan countries to find out whether investment climate (like number of outages, share of firms with bank loans and others) fosters employment growth and whether there exist some similarities among the countries. They concluded that average firm level employment growth rate is quite similar in spite of differences in the quality of investment climate. Although this study offers useful insights from disaggregated data it fails to test directly the effect of investment and in particular private investment on employment which is the focus of this paper. Other factors have also been generally linked with employment generation, in addition to investment. From surveys, Afram and Del Pero (2012) report that even though Nepal has recorded growth in certain niche sectors and private sector employment is increasing (by almost 4 percent a year between FY2005/06 and FY2007/08) constraints (political instability, poor infrastructure, poor labor relations, poor access to finance, and declining exports) to the investment climate are hindering this progress. Aterido, Hallward-Driemeier and Pagés (2007) reveal that business environment that does not support access to finance and business regulation reduce the employment growth of all firms but micro and small firms suffer the most. Also, corruption and poor access to infrastructure are detrimental to employment growth of medium size and large firms. These conclusions were based on World Bank Enterprise Survey (WES) data from about 70,000 enterprises in 107 countries. In other words, institutional and structural variables play a key role in labour demand analysis Pierluigi and Roma (2008) Heintz (2001) postulated that increases in political instability explain the largest portion of the decline in the rate of investment in South Africa over this period. In that study, econometric estimates showed significant negative effects of higher average product wages and greater political unrest on the labor-capital ratio. Also, among South African manufacturing firms, Behar and Edwards (2006), report that trade liberalization and technological change have affected the skill structure of employment. They explain that export orientation, raw materials imports, training, investment in computers and firm age are positively associated with the skill intensity of production. Some research also link employment with human capital development (Pryor and Schaffer, 1999; Card, 1999 and; Wolman et al., 2008). In Nigeria, Aromolaran (2004) postulates that even though private returns to schooling associated with levels of educational attainment for wage and self employed workers are low at primary and secondary level, they are substantial at post-secondary education level. But then discrimination and technology seem to reduce the magnitude of this effect. For instance, Bertrand and Mullainathon (2003) argue that whites experience higher return to more resume credentials than blacks. Also, about a decade ago, Autor, Levy and Murnane (2003) asserted that technology is taking over routine jobs. A position Manning (2013) does not only support but argues that is the main reason for job polarisation and its associated inequality (Goos and Manning, 2007). Baldwin (1995) concluded that the employment benefits of increased in exports far outweigh the employment-displacing effects of increased imports even though the study could not conclude on the employment effects of foreign direct investment (FDI). Nickell (2010)

11

argues that even though unemployment is falling in Europe, in the credit crunch recovery period, if there is no rise in GDP growth, this fall may not have any significant impact. Apparently, what the above reviewed literature suggests is that in addition to factors such as real factor cost, human capital, governance, aggregate demand, economic growth and trade openness, investment and in particular private investment has a potential impact on the level of employment. However, in Africa, empirical knowledge of this relationship is yet to be known. Hence, the need for this study. 3.0 Methodology 3.1 Study sample The study included data from 48 countries in Sub-Saharan Africa excluding South Sudan. The exclusion of South Sudan was basically based on lack of data. All these countries are studied over a 20 year period, from 1990 to 2009. 3.2 Data

All the data were taken from the online edition of the African development index of the World Bank except that of Trade openness and Polconiii. The variable for trade openness was taken from UNCTAD but that of Political Discretion (Pol) is an index built by Henisz (2010). All the variables, except governance (Gov), political discretion (Pol), human capital and Agricultural Productivity Index (API), are presented in their natural log form in order to control for heteroskedasticity.

The Gov variable is measured as an index constructed (using the Principal Component Analysis - PCA) from the global governance indicators published by the World Bank. The procedure for the construction of the PCA is shown in appendix 1.

3.3 The Neoclassical Labour Demand Model The demand for labour is derived from the demand for output. Thus the demand for labour is influenced by cost of labour and output. The neoclassical labour demand model begins on the premise of the firms profit function:

where is profit is nominal wage is labour and is cost of capital and capital respectively. It is assumed that profit is maximised subject to and also when change in profit relative to changes in employment and capital are equal to zero. Thus, it is a theory of distribution which specifies how output is shared among the various factors of production. The neoclassical model depicts that labour demand curve slopes downward whilst the labour demand is also shown to be wage elastic. According to the model, real wage is of more importance than nominal wage because it caters for the effect of price changes, facilitates comparison of wages for two countries and/or comparison of wages for one country but overtime. Given that nominal wage = = GHS per unit of and product price ( =GHS per unit of , real wage = = unit of per unit of . This is equivalent to the marginal product of labour (MPL) = . Thus, the model does not only specify that labour should be hired up to the point where marginal revenue is just equal to marginal cost, as a result of diminishing returns but also real wages can only rise (when prices are constant) when real labour productivity has risen.

12

According to the neo-classical model of labour demand, in the short run the only factor that influences labour demand is the real wage because capital is assumed to be constant. On the other hand, in the long run labour demand is dependent on both cost of capital and the real wage rate, in addition to output. Thus the long run neoclassical labour demand model could be written as follows:

YrwEELT ,, (1) Where: is long term labour demand; w is real wage rate; r is cost of capital and Y is output. Next, we explain output, from a Cobb-Douglas (1928) production function in which private investment is explicitly incorporated.

1,, pgtt KKLA (2) Where is output, L is human capital, is public investment, pK is private investment, is the efficiency of production, , and 1 are the human capital, public investment and private investment shares respectively. When we put equation (2) into equation (1) we get

1,, pgtLT KKLArwE (3) When the natural logs are taken, we get:

pgLT KKLArwE ln)1(lnlnlnlnlnln (4) Equation 4 could be written as:

pgLT KaKaLarawaaE lnlnlnlnln 543210 (5) Where:

is the natural log of employment generation is the natural log of labour cost measured as real wage cost

is the natural log of cost of capital measured as real interest rate is the natural log of human capital

is the natural log of Capital (public investment or investment by government) and; is the measure of private investment.

Equation (5) says that when output is explained as human capital, public investment and private investment, additions to private investment will enhance output, by a factor 1 , and hence labour demand. The nature of the data allows for the use of panel data methodology for the analysis. Panel data methodology has the advantage of not only allowing researchers to undertake cross-sectional observations over several time periods, but also control for individual heterogeneity due to hidden factors, which, if neglected in time-series or cross-section estimations leads to biased results (Baltagi, 1995). The general form of the panel data model can be specified as:

13

(6)

Where the subscript i denotes the cross-sectional dimension (equal to 1……48), and t represents the time-series dimension (1 to 20 years). Yit, represents the dependent variables in the model, which are total, male and female employment. X contains the set of explanatory variables in the estimation model. a is the constant and ß represents the coefficients.

where is an unobserved individual specific effect, is an unobserved time specific effect and is a zero mean random disturbance with a variance of . The error components of variance follow a two way analysis of variance (ANOVA). If and denote fixed parameters to be estimated, this model is known as the fixed effects. In this case the ‘s are assumed independent of the ‘s for all i and t. On the other hand if and are random variables with zero means and constant variances and , this model is known as

the random effects model . The random effects model also adds and

showing that the two error components are independent from each other (Baltagi, 2008 and Hsiao, 2003). The study used the random effects model was used because it addresses the problems of variable omission bias and the use of unbalanced panels with unequally spaced data, which is the case with the SSA data used for the study (Baltagi and Wu,1999 and Asiedu, 2004). Also the Hausman (1978) test influenced the choice of random effects estimation. In the Hausman test the null hypothesis is that the preferred model is random-effects. In other words, the unique errors ( are not correlated with the regressors (Greene, 2008). In most of the estimations, the chi2 of the Hausman (1978) test was negative making it difficult to choose between random effects and fixed effects models. In such a situation the researcher could conclude that the random effects model is okay (Greene, 2008: 209 and Park, 2011). In order to ensure that changes in employment is solely due to changes in private investment and also to assess the effect of some other variables on employment demand we included proxies for public sector investment, human capital, trade openness, governance, economic growth, final consumption expenditure and productivity of the agricultural sector. Also we wanted to check whether private investment influenced the demand for total labour the same way as male labour demand, female labour demand, total youth, male and female youth labour demands. This resulted in estimating six main models. The following expanded six main models were thus estimated: lnEMPTOTit = α +β1lnRWRit + β2lnRIRit + β3HDIit + β4POL3it + β5GOVit +β6lnTOPENit +

β7lnGPINVit+ β8lnPRINVit + β9 lnGROit + β10 lnFCEGDPit + β11 APIit + εit (7)

lnEMPMALit = α +β1lnRWRit + β2lnRIRit + β3HDIit + β4POL3it + β5GOVit +β6lnTOPENit +

β7lnGPINVit+ β8lnPRINVit + β9 lnGROit + β10 lnFCEGDPit + β11 APIit + εit (8) lnEMPFEMit = α +β1lnRWRit + β2lnRIRit + β3HDIit + β4POL3it + β5GOVit +β6lnTOPENit +

β7lnGPINVit+ β8lnPRINVit + β9 lnGROit + β10 lnFCEGDPit + β11 APIit + εit (9)

14

lnEMPTOTYit = α +β1lnRWRit + β2lnRIRit + β3HDIit + β4POL3it + β5GOVit +β6lnTOPENit +

β7lnGPINVit+ β8lnPRINVit + β9 lnGROit + β10 lnFCEGDPit + β11 APIit + εit

(10) lnEMPMALYit = α +β1lnRWRit + β2lnRIRit + β3HDIit + β4POL3it + β5GOVit +β6lnTOPENit +

β7lnGPINVit+ β8lnPRINVit + β9 lnGROit + β10 lnFCEGDPit + β11 APIit + εit

(11) lnEMPFEMYit = α +β1lnRWRit + β2lnRIRit + β3HDIit + β4POL3it + β5GOVit +β6lnTOPENit +

β7lnGPINVit+ β8lnPRINVit + β9 lnGROit + β10 lnFCEGDPit + β11 APIit + εit

(12) Dynamic Labour Demand In addition, we included three more models to estimate the dynamic nature of labour and whether adjustment cost influenced labour demand. The first two models were estimated with the total youth labour demand and the male labour demand as dependent variables. These two dependent variables were chosen because, at their levels, real wage cost and cost of capital did not have any effect on youth employment and male employment. So the study included one year lag of the real wage and capital costs variables in addition to the lags of their dependent variables in the dynamic models. The assumption is that real wage cost and cost of capital sometimes have delayed effect on employment because of adjustment cost. The third dynamic model that was estimated used the female labour demand as the dependent variable because this is the only model, in the six random effects models that were estimated, that did not show a decisive effect of private investment on female labour demand. The estimation of the dynamic model of the female labour demand is based on the assumption that investment and employment have lagged endogeneous regressors as well as unobserved country fixed-effects which are correlated with the regressors.

In all the dynamic models, the Arellano and Bond (1991) estimation technique was used. It is an instrumental variable (IV) estimator that accounts for correlated fixed effects and endogenous regressors (Asiedu and Gyimah-Brempong, 2008). Subsequent to the AB estimation the Sargan and autocorrelation tests were applied to identify whether the models were well specified. The Sargan test for over-identifying restrictions is used to determine if the instruments are suitable. The null hypothesis states that “the instruments as a group are exogenous”. Consequently, a higher p-value is preferred. The null hypothesis of no autocorrelation is applied to the differenced residuals (Mileva, 2007). Sargan test results and results for AR (1) and AR (2) test reported in Table 8 shows that the model is well specified.

The dynamic models are shown below.

lnEMPTOTYit = α + β1lnEMPTOTYit-1 +β2lnRWRit + β3lnRWRit-1 + β4lnRIRit +β5lnRIRit-1+ β6HDIit + β7POL3it + β8GOVit +β9lnTOPENit + β10lnGPINVit+ β11lnPRINVit + β12 lnGROit + β13 lnFCEGDPit + β14 APIit + εit (13)

15

lnEMPMALit = α + β1lnMPMALit-1 + β2lnRWRit + β3lnRWRit-1 + β4lnRIRit +β5lnRIRit-1+ β6HDIit + β7POL3it + β8GOVit +β9lnTOPENit + β10lnGPINVit+ β11lnPRINVit + β12 lnGROit + β13 lnFCEGDPit + β14 APIit + εit (14)

lnEMPFEMit = α + β1lnEMPFEMit-1 +β2lnRWRit + β3lnRWRit-1 + β4lnRIRit +β5lnRIRit-1+ β6HDIit + β7POL3it + β8GOVit +β9lnTOPENit + β10lnGPINVit+ β11lnPRINVit + β12lnPRINVit-1 + β13 lnGROit + β14 lnFCEGDPit + β15 APIit + εit (15)

The definition of the variables used in the study and their expected signs are provided in Table 2 below Table 2: Definition of variables (proxies) and Expected signs VARIABLE DEFINITION THEORIES EXPECTED

SIGN EMPTOT Total Employment (Dependent Variable)

= Total Employment to Total Population ratio is the proportion of a country's population that is employed. Ages 15 and older are generally considered the working-age population. This is calculated for country i in time t;

Neoclassical Labour Demand

EMPMAL Male Employment (Dependent Variable) = Male Employment to Male population ratio is the proportion of a country's population that is employed. Ages 15 and older are generally considered the working-age population. This is calculated for country i in time t;

Neoclassical Labour Demand

EMPFEM Female Employment (Dependent Variable) = Female Employment to Female Population ratio is the proportion of a country's Female population that is employed. i.e. Percentage of total employment that is female for country i in time t. Ages 15 and older are generally considered the working-age population.

Neoclassical Labour Demand

16

EMPTOTY Employment to population ratio is the proportion of a country's population that is employed. Proportion of total youth employed for country I in time t. Ages 15-24 are generally considered the youth population.

EMPMALY Employment to population ratio is the proportion of a country's population that is employed. Proportion of male youth employed for country I in time t. Ages 15-24 are generally considered the youth population.

EMPFEMY Employment to population ratio is the proportion of a country's population that is employed. Proportion of female youth employed for country I in time t. Ages 15-24 are generally considered the youth population.

RWR Real Wage Rate = Nominal Wage Rate/Consumer Price Index for country i in time t; Nominal Wage Rate is Compensation of employees as a percentage of total expenses for country i in time t; Compensation of employees consists of all payments in cash, as well as in kind (such as food and housing), to employees in return for services rendered, and government contributions to social insurance schemes such as social security and pensions that provide benefits to employees.

Neoclassical Labour Demand

Negative

17

RIR Real Interest Rate (independent Variable) = is the year end real interest rate of country i in time t;

Neoclassical Labour Demand

Negative

HDI Human Capital Index = Measures 2 indicators (a) Health and Welfare and (b) education. It is based on Ibrahim Index measures reported by the World Bank for country i in time t;

Neoclassical Labour Demand

Positive

POL3(Polconiii)

Political Discretion/Constraint = It is measured as the level of political discretion or constraint and ranges from 1 (political discretion) to 0 (political constraint) of country i in time t based on Henisz (2010);

Governance positive

GOV Country Governance Index (1): Is an index constructed using principal component analysis from six global governance indicators provided by the world bank. The index is constructed for country I in time t

Stewardship Theory

TOPEN Trade openness = This shows exports, imports and sum/average of exports and imports as percentage of nominal gross domestic product (GDP) for country i in time t. The indicators are calculated for trade in goods, trade in services and total trade in goods and services. The data is taken from UNCTAD Database

Structural Adjustment

Indeterminate

18

GPINV Gross Public Investment =

Gross public investment (see definition below) as a percentage of GDP (%). Public sectors’ gross domestic fixed investment (gross fixed capital formation) comprises all additions to the stocks of fixed assets (purchases and own-account capital formation), less any sales of second-hand and scrapped fixed assets measured at constant prices, done by government units and non-financial public enterprises. Most outlays by government on military equipment are excluded. It is calculated for country i in time t;

Neoclassical Labour Demand

Positive

PRINV Private Investment (Gross Fixed Capital Formation by the Private Sector) = investment output ratio and is computed as the ratio of private investment to GDP of country i in time t. Private investment covers gross outlays by the private sector (including private non-profit agencies) on additions to its fixed domestic assets.

Neoclassical Labour Demand

Positive

GRO Annual percentage growth rate of GDP per capita based on constant local currency for country i in time t. GDP per capita is gross domestic product divided by midyear population. GDP at purchaser's prices is the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products. It is calculated without making deductions for depreciation of fabricated assets or for depletion and degradation of natural resources.

Growth Positive

19

FCEGDP Final consumption expenditure (formerly total consumption) is the sum of household final consumption expenditure (private consumption) and general government final consumption expenditure (general government consumption) scaled by GDP. This estimate includes any statistical discrepancy in the use of resources relative to the supply of resources for country i in time t.

Indeterminate

API Agriculture Production Index =

The FAO indices of agricultural production show the relative level of the aggregate volume of agricultural production for each year in comparison with the base period 1999-2001. They are based on the sum of price-weighted quantities of different agricultural commodities produced after deductions of quantities used as seed and feed weighted in a similar manner. The resulting aggregate represents, therefore, disposable production for any use except as seed and feed. This is calculated for country i in time t;

Positive

Ɛ The error term

4.0 Discussion of Empirical Results 4.1 Descriptive Statistics Table 3 gives the descriptive statistics of the variables used in the study. Total employment level among the working population in Africa is about 64.13%. Expectedly more men (72.26%) are engaged in employment than women (56.29%) probably because of the traditional role of men in most African cultures. The dispersion among female employment is quite worrying. The records showed that some countries recorded as low as 12.7% while others as high as 88.2%. Mauritania recorded the minimum total female, female youth and total employment levels whiles Rwanda achieved the maximum total female, female youth,

20

and total employment levels. Meanwhile, the maximum levels of total investment, public investment and private investment were recorded by Equatorial Guinea (see Appendix 5) During the study period, employers spent about 35% of their total expenses on their workforce for engaging their services, with the lowest and highest rates being 10.2% and 60.6% respectively. Meanwhile real interest rate averaged at 10.8%. Private investment (12.6%), over the two decades of study, was greater than the level of governments’ (7.5% of GDP) involvement in investment activities. Privatization of state-owned enterprises and the proliferation of non-governmental organisation could be contributing factors. But what is also clear is that the continent is a consuming continent. This is reflected in the fact that about 91% of the region’s GDP is covered by final consumption expenditure. This is also reflected in the size of trade openness. Economic growth, during the period was low (4%) with some countries retrogressing (-50%). This notwithstanding, the productivity of the agricultural sector appeared to be relatively representative. Table 3: Descriptive Statistics Var. Obs. Mean Std Dev Min Max Emptot 855 64.13345 12.62545 31.8 88.3 Empmal 855 72.25497 10.24244 44.1 88.6 Empfem 855 56.29135 17.46347 12.7 88.2 Emptoty 855 47.16795 16.03483 10.5 80 Empmaly 855 51.33673 16.3792 13.9 79.7 Empfemy 855 43.01439 17.7237 6.1 81.1 Nwr 262 35.5873 10.96368 10.1795 60.6036 rir 641 10.84186 27.76049 -96.8698 508.741 hdw 470 49.61005 14.89904 10.3805 89.4437 polconiii 422 0.3172512 0.1524683 0 0.73 ps 532 0.4709884 18.11216 -33.6954 31.6019 tradeopen 838 31.4506 21.24236 2.68738 140.576 GPINV 841 7.407808 4.825831 0.100101 42.9755 PRINV 840 12.75484 9.776949 -2.64039 112.352 GRO 925 1.423028 7.758877 -50.2904 92.586 FCEGDP 865 91.06987 18.75364 13.1179 187.537 API 954 88.52479 19.37031 37.67 208.04

4.2 Multicollinearity

In order to test for the presence of multicollinearity among the regressors, two main tests were conducted. The correlation among the variables was estimated just as their variance inflation factors (VIF). The results, as indicated in Table 4 and 5 show that the presence of multicollinearity is minimal. This is reflected in the low correlation values and a low mean VIF of 2.52. Multicollinearity is deemed to be high if VIF is greater than 5 (as a common rule of thumb) and according to Kutner (2004), VIF of 10 should be the cut off.

Table 4: Variance inflation Test VIF 1/VIF lntradeopen 3.82 0.262084

21

lnPrinv 3.66 0.273568 Hdw 3.64 0.273568 lnFcegdp 3.50 0.285597 lnNerwr 2.46 0.406079 lnRir 2.22 0.450552 Gov 2.18 0.458517 lnGpinv 1.83 0.546215 Pol3 1.69 0.593374 lnApi 1.56 0.642085 lnGro 1.16 0.864146 Mean VIF 2.52

22

Table 5: Correlation Matrix

lnEmptot lnEmpmal lnEmpfem lnEmptoty lnEmpmaly lnEmpfemy lnRwr lnRir Hdw Pol3 Ps lntradeopenlnEmptot 1.0000 lnEmpmal 0.8357*** 1.00000 lnEmpfem 0.9321*** 0.5878*** 1.0000 lnEmptoty 0.9143*** 0.9203*** 0.7615*** 1.0000 lnEmpmaly 0.7690*** 0.9426*** 0.5375*** 0.9477*** 1.0000 lnEmpfemy 0.9634*** 0.7984*** 0.9137*** 0.9460*** 0.7960*** 1.0000 lnRwr 0.1229* 0.0615 0.1542** 0.0349 -0.0556 0.1138* 1.0000 lnRir 0.1860*** 0.1510*** 0.1730*** 0.1640*** 0.1292*** 0.1777*** 0.1776** 1.0000

Hdw -0.2341***

-0.2418***

-0.1787*** -0.2953 -0.2609*** -0.2911*** -0.1901** -0.1085* 1.0000

Pol3 0.0149 0.1537*** -0.0642 0.0617 0.1034** 0.0107 -0.0669 0.0905 0.0026 1.0000 Ps 0.0606 0.0767* 0.0472 0.1338*** 0.1508*** 0.1049** -0.0054 -0.0058 -0.0369 -0.0678 1.0000

lntradeopen -0.2883***

-0.3323***

-0.1873***

-0.3161*** -0.2942*** -0.2847*** 0.1046 0.0852 0.3054*** 0.0203 0.0855* 1.0000

lnGpinv 0.1176*** 0.1284*** 0.1183*** 0.1977*** 0.2180*** 0.1683*** -0.2871***

-0.1173*** 0.1433*** -0.0467 0.0295 -0.0671*

lnPrinv -0.1936***

-0.1639***

-0.1624***

-0.1645*** -0.1012*** -0.1966***

-0.3057*** -0.0162 0.3731***

-0.1307*** 0.0147 0.3617***

lnGro 0.1015** 0.1066*** 0.0726* 0.1325*** 0.1372*** 0.1094*** 0.0485 -0.0471 -0.0138 -0.018 0.0132 0.0903**

lnFcegdp 0.0459 0.0977*** 0.0085 0.1751*** 0.1787*** 0.1404*** 0.0125 0.0087 -0.1162** 0.0919* 0.0756* -0.3955***

lnapi -0.0393 -0.0944*** 0.0107

-0.0902*** -0.0980*** -0.0623* 0.1432 0.0215 0.1850*** -0.0727 0.0014 0.3681***

23

4.3 Discussion of Regression Results The results, from the random effects model, show that private investment together with public investment, economic growth, governance and to some extent real wage cost, are among the key factors that increase total employment in Africa. On the other hand, employment in Sub Saharan Africa is attenuated by trade openness, human capital development, increase in agricultural productivity and to some extent final consumption expenditure. The results indicate a strong positive relationship between private investment and total, male, and youth employment demand. Apparently private investors compliment government efforts in solving the unemployment problem that has bedeviled Africa. The channel of this effect could be direct or indirect. Direct, by engaging the populace to facilitate their productive activities and indirect, by amplifying the labour demand of other organisations through linkages with other sectors, be it public or private. Manufacturing and service industries are better placed in influencing employment on the continent especially through the indirect channel than the extractive industry because the extractive industry does not only have less linkages with other sectors of the economy but also because it has huge capital requirement. For instance, the upsurge of private investment in the telecommunication industry, created the platform for growth in entrepreneurial activities among small and medium scale enterprises in facilitating the distribution of such services. Similarly, public investment is a major source of employment to Africans. This is possible due to the fact that investments by the state are generally not for profit motive. Consequently, these results confirm the exceptional role governments play in employment generation on the African continent. This result does not support the assertion that investment on the continent is not productive, at least from the point of view of employment generation. But then seem to sit well with the strand of literature that argues for more investment to facilitate the development of the continent. This notwithstanding, public investment appears to have a stronger impact on employment generation than private investment. Against the prediction of the neoclassical labour demand, wage rate increases advances the purchasing power ability of the working class thus spurring on productivity and employment. This seems to suggest that most SSA countries are depressed and that an increase in wage rate would be beneficial to employment demand than a reduction. Also, economies that are politically stable do not only allow their citizens to enjoy their wealth peacefully but are also able to increase the confidence of industries of the safety of their investments. As a result, political stability may operate indirectly either through the consumption channel of investment channels. Trade openness does not favour employment in SSA mainly because the continent is a net importer. Even though the continent is endowed with a lot of primary resources, its weak manufacturing sector means that most of these resources are exported at their raw stages at less competitive prices than their eventual final products. Ironically, the continent serves as a major market for these final products, worsening our net export position. Similarly, increase in the consumption of goods and services does not translate into increase in employment by rather harms it. Probably because, most of the goods and services consumed in the region are imported. Thus an increase in consumption of these goods and services increases the employment demand of the manufacturing country and not the consuming country. Surprisingly, human capital measured as human development index consistently and significantly shows a negative relationship with employment demand. This could probably be as a result of the fact that, given the developmental stage of the continent, there do not exist

24

enough job openings for highly skilled workers. Also, the negative relationship between human capital and employment seem to suggest that the SSA economy has not been expanding large enough to accommodate the kind of human capital that exist in the region. Consequently, any improvement in human capital which increases the productivity per worker means that less people will have to be engaged, thereby harming employment. This position is similar to the negative effect technological advancement has on employment. Moreover, the results show a positive and significant relationship between agricultural productivity index and employment. Even though this result is counter intuitive, especially given that the sector major source of employment for SSAs, it is not surprising. Most of the citizens engaged in the agric sector are in the area of persistence farming. Thus if the expansion in the Agric sector is largely fuelled by mechanization, then it would harm labour demand. The results for male and female labour demands do not differ significantly from total labour demand except for the fact that private investment does not support female (excluding female youth) employment. The reason is because most of the largest external financial flow (FDI) into the African continent go into hunt for resources (AfDB,OECD,UNECA, 2012) and employment in these areas is male dominated. The results are also robust for youth employment except that results for real wage and final consumption expenditure are not statistically significant. Table 6: Regression Results for model 7, 8 and 9 ALL Total Male Female lnRwr 0.0652385*** 0.027989 0.1185152*** (0.21292) (0.0199707) (0.0318429) lnRir 0.002591 0.0165279 -0.0179494 (0.0186182) (0.0174628) (0.0278441)

Hdi -0.0106018*** -0.0054872*** -0.0172796***

(0.0012939) (0.0012136) (0.0019351) Pol3 0.3444724*** 0.29749*** 0.3786572*** (0.0572351) (0.0536832) (0.085597) Gov 0.0067629*** 0.0056843*** 0.0081529*** (0.0005307) (0.0004978) (0.0007937)

lnTradeopen -0.2028684*** -0.0923008*** -0.3439824***

(0.0286867) (0.0269065) (0.042902) lnGpinv 0.1520607*** 0.1490745*** 0.16010904*** (0.027583) (0.0258713) (0.0412513) lnPrinv 0.0957582** 0.0875658** 0.1081035 (0.0457667) (0.0429266) (0.0684457) lnGro 0.0148426** 0.0127689** 0.0168818* (0.0067084) (0.0062921) (0.0100327) lnFcegdp -0.5498374** -0.1954293 -1.075349*** (0.2614354) (0.2452114) (0.3909854)

api -0.440721*** -0.4697842*** -0.3913844**

25

(0.1259751) (0.1181574) (0.1884) Constant 9.294384*** 7.247534*** 12.20781*** (1.238538) (1.161678) (1.852275) R2 = 0.9697 0.9547 0.9651 Wald Chi2 (11) = 734.83 484.24 635.22 Prob. > Chi2 = 0.0000 0.0000 0.0000 Haus. Test Chi2(11) -141.07 -143.40 -25.99 *** = 1%, ** =5% and * = 10% Standard Errors in Parenthesis

Table 7: Regression Results for models 10, 11 and 12 YOUTH Total Male Female lnRwr 0.0033037 0.0045292 0.0107302 (0.0603413) (0.0610077) (0.0600066) lnRir 0.0042437 0.0061034 0.0026891 (0.0527637) (0.0533465) (0.0524711) Hdi -0.0212348*** -0.0150644*** -0.028652*** (0.0036669) (0.0037074) (0.0036465) Pol3 0.7006179*** 0.552695*** 0.8497684*** (0.1622036) (0.1639951) (0.1613041) Gov 0.0130406*** 0.0131871*** 0.0135369*** (0.001504) (0.0015206) (0.0014957) lnTradeopen -0.1571773** -0.0964921 -0.2365264*** (0.0812979) (0.0821958) (0.0808471) lnGpinv 0.4140286*** 0.4150941*** 0.4138402*** (0.07817) (0.0790334) (0.0777365) lnPrinv 0.2665101** 0.2635558** 0.2649171** (0.1297024) (0.131135) (0.1289832) lnGro 0.045581** 0.0386213** 0.0516013*** (0.0190116) (0.0192216) (0.0189062) lnFcegdp 0.0852108 -0.055203 0.0059756 (0.7409051) (0.7490882) (0.7367965) api -1.173396*** -1.123644*** -1.251793*** (0.3570121) (0.3609552) (0.3550323) Constant 8.653926*** 8.614928** 9.942307*** (3.510004) (3.548771) (3.49054) R2 = 0.9393 0.9241 0.9556 Wald Chi2 (11) = 355.97 280.03 494.77 Prob. > Chi2 = 0.0000 0.0000 0.0000 Haus. Test Chi2(11) -256.81 -12.38 -203.64 *** = 1%, ** =5% and * = 10% Standard Errors in Parenthesis

26

Dynamic Labour Demand The dynamic models estimated in table 8 shows strongly that employment levels in the previous year informs employment levels in the current year at 1% significant level. This implies that when people are engaged they do not easily lose their employment. In the same way those who are unemployed find it difficult securing employment. This suggests the presence of adjustment cost. The neoclassical labour demand model appears to be properly situated in the dynamic labour demand model. In all the estimated models, real wage rate and real interest rates negatively influence labour demand- but at their levels and not their lags. The implication being that the adjustment process of wage cost and real interest cost does not travel beyond one year, in SSA. Within the dynamic framework, even though private investment positively influences labour demand for total youth, male and female labour demand, it is only significant for total youth employment. In fact, in the case of female labour demand, previous years’ private investment reduces current employment levels. The results from the dynamic model are generally consistent with the random effects model in terms of the effect of private investment, agricultural productivity, trade openness and political stability on employment. But they differ in terms of results for human capital, public investment and growth. Thus, from the dynamic model, the study concludes that: 1) private investment is a significant source of youth employment in Sub-Saharan Africa; 2) adjustment cost influences the effect of change in factor cost on labour demand and that 3) support for the neoclassical labour demand model is evident when cost of adjustment (even though adjustment process takes less than a year) is factored into the model Table 8: Dynamic Labour Demand Models 13, 14 and 15 lnEmptoty lnEmpmal lnEmpfem lnEmptoty t-1 1.309833*** (0.044429) lnEmpmal t-1 0.8399709*** (0.09465) lnEmpfem t-1 0.7959103*** (0.0709719) lnnerwr -0.028244** -0.0142057* -0.0134222 (0.0128658) (0.0075416) (0.0087025) lnnerwr t-1 -0.0070162 -0.0039998 -0.002644 (0.0053683) (0.0036389) (0.0032055) lnRir -0.0189666*** -0.0037597*** -0.0035355** (0.0020131) (0.0013078) (0.0016228) lnRirt-1 0.0082298*** 0.002745*** 0.0035116 (0.0026551) (0.0010827) (0.0013643) hdi 0.0024803*** 0.000426 0.0002464 (0.0008188) (0.0005317) (0.00044464) lnGpinv 0.0054419 -0.0053764* -0.0089551*** (0.0054716) (0.0029376) (0.003054) lnPrinv 0.0181513*** 0.0003923 0.0012941 (0.0060084) (0.0031851) (0.0025572) lnPrinv t-1 -0.010312***

27

(0.003346) Gov 0.0002854 -0.0003149 -0.0002933 (0.0005964) (0.0003126) (0.0003352) Pol3 0.0991032*** 0.0103181 0.0105125 (0.0181961) (0.0127497) (0.0122475) lntradeopen -0.0567221** -0.0001549 --0.0130965 (0.0285097) (0.0149592) (0.0116454) api -0.0012696*** -0.0001549 0.0001623 (0.0004404) (0.000187) (0.0002082) lnfcegdp 0.0329964 0.0228602 0.0392215** (0.0630091) (0.0267312) (0.0200533) lnGro -0.0107681*** -0.0038523*** -0.0025357** (0.0028868) (0.0013619) (0.0011572) Constant -1.251101*** 0.6206504 0.7271406** (0.4929269) (0.4806525) (0.3376218) Wald Chi2(6) 147125.95 170000000 93499.01 Prob>Chi2 0.0000 0.0000 0.0000 Sargan Test:

Chi2 (8) 7.224748 8.153962 7.040145

Prob. 0.5126 0.4186 0.4247 Autocorrelation

1 z(Prob.) 1.0352(0.3006) 0.81206(0.4168) 0.78911 (0.4301)

2 z(Prob.) 0.74208 (0.4580) 0.92519(0.3549) 1.0585 (0.2898)

*** = 1%, ** =5% and * = 10% robust Standard errors in parenthesis 5.0 Conclusion We set out with the basic objective of assessing whether employment generation is part of the benefits that Sub Saharan African economies can get from private investment, which some consider not to be enough (Dinh et al (2012)) and others unproductive (Devarajan, et al 2002). Data was taken from the World Bank, UNCTAD and Henisz (2005) covering 48 Sub-Saharan African countries over a period of 20years (1990-2009). We estimated a derived neoclassical model that allows for the inclusion of private investment, labour cost, real interest rate, human capital and investment by the public sector. The model also controls for governance, trade openness, economic growth, final consumption expenditure and agricultural productivity. Within the framework of panel methodology, the model was then estimated with the random-effects model and subsequently the dynamic labour demand model was used to cater for the effect of adjustment process and cost. The results from the random effects model suggest strongly that private investment is beneficial to SSA in terms of employment generation especially for the entire working population and male population. Unfortunately, however, women seem not to benefit substantially from increases in private sector investment. Women on the continent benefit more when investment is from government than private institutions. Also, public investment, economic growth, governance and to some extent real labour cost are catalyst for improving employment on the African continent. Nonetheless, trade openness, final consumption

28

expenditure, surprisingly human capital development and agricultural productivity are harmful to employment generation. Results from the dynamic model lend strong support for the neoclassical demand –after accounting for adjustment process and cost. But then the adjustment process does not go beyond one year. Thus while the relationship between real wage cost, cost of capital and labour demand are significantly negative, that of private investment and labour demand is significantly positive. In effect, this result does not only offer support for the neoclassical labour demand model but also stress that private investment offers employment to Sub Saharan Africa. Consequently, the SSA region should intensify measures to attract private investment to more productive areas especially manufacturing, motivate the private sector through tax incentives, improve on the judicious use of public investment through checking corruption and ensuring value for money investments, promote exports, and embark on policies that grow the economy. In addition, Conscious effort should be made to assess the impact of investment on the economy. This impact assessment should include employment impact of investment assessment which should be handled by a body independent of that which granted the permit for investment in order to ensure objectivity. References Addison, J., L. Bellmann, T. Schank, and Paulino (2008). “The Demand for Labor:An Analysis Using Matched Employer - Employee Data from the Germna LIAB. Will the High Unskilled Worker Own-Wage Elasticity Please Stand Up?” Journal of Labor Research 29 (2), 114 -137. Afram, G. G. and Del Pero, A. S. (2012) “Nepal’s Investment Climate: Leveraging the Private Sector for Job Creation and Growth.” Directions in Development (Private Sector development) World Bank Report 67525 DOI: 10.1596/978-0-8213-9465-6abi Andrews, Martyn & Nickell, Stephen, (1982). “Unemployment in the United Kingdom since the War,” Review of Economic Studies, Wiley Blackwell, vol. 49(5), pages 731-59, Special I.

Aromolaran, Adebayo, B. (2004). “Wage Returns to Schooling in Nigeria” African Development Review, Vol. 16, No.3, pp. 433 – 455. Asiedu, E. (2004) “The Determinants of Employment of Affiliates of US Multinational Enterprises in Africa” Development Policy Review, 2004, 22 (4): 371-379. Asiedu, Elizabeth and Gyimah-Brempong, Kwabena (2008) “The Effect of the Liberalization of Investment Policies on Employment and Investment of Multinational Corporations in Africa.” African Development Review, Vol. 20, No.1, pp. 49-66. Assaad, Ragui; El-Hamidi, Fatma and Ahmed, Akhter U. (2000) “The Determinants of Employment Status In Egypt” International Food Policy Research Institute - Food Consumption and Nutrition Division (IFPRI- FCND) Discussion Paper No. 88 pp. 1-107.

29

Aterido, R., Hallward-Driemeier , M. and Pagés, C. (2007) “Investment Climate and Employment Growth: The Impact of Access to Finance, Corruption and Regulations Accross Firms” Forschungsinstitut zur Zukunft der Arbeit. Institute for the Study of Labor Discussion Paper IZA DP No. 3138.

Aterido, R. and Hallward-Driemeier, M. (2010). “The Impact of the Investment Climate on

Employment Growth: Does Sub-Saharan Africa Mirror Other Low-Income Regions?” Policy

Research Working Paper - WPS5218, Developmental Research Group, World Bank Autor, David, Levy, Frank and Murnane, Richard (2003). ‘The Skill Content of Recent Technological Change: An Empirical Exploration’, Quarterly Journal of Economics 118(4): 1279-1333. Baldwin, R. E. (1995), “The Effects of Trade and Foreign Direct Investment on Employment and Relative Wages”, OECD Jobs Study Working Papers, No. 4, OECD Publishing. http://dx.doi.org/10.1787/888157653682

Baltagi, Badi H., (2008). “Econometric Analysis of Panel Data,” John Wiley & Sons, Ltd.

Barba Navaretti, G., A. Turrini, and D. Checchi (2003). “Adjusting Labor Demand: Multinational Versus National Firms: A Cross-European Analysis.” Journal of the European Economic Association 1 (2-3), 708 - 719. Barrell, R., N. Pain and G. Young (1996). “A Cross-Country Comparison of the Demand for Labour in Europe.” Weltwirtschaftliches Archiv 132, 638-650.

Barro, Robert J. and Grossman, Herschel I. (1971). “A General Disequilibrium Model of Income and Employment” The American Economic Review Vol. 61, No. 1, pp. 82-93 Article Stable URL: http://www.jstor.org/stable/1910543

Beltagi, B. H., (1995). “Econometric Analysis of Panel Data.” Wiley, Chichester Bentolila, S. – Gilles, St. P. (1992). “The Macroeconomic Impact of Flexible Labor Contracts, with an Application to Spain.” European Economic Review, June, Vol. 36. No. 5. pp. 1013–1053. Bentolila, Samuel and Bertola, Giuseppe (1990). “Firing Cost and Labour Demand: How Bad is Euroclerosis?,” Review of Economic studies 57, 381 – 402.

Berman, E., J. Bound, and Z. Griliches (1994). “Changes in the demand for skilled labor within US manufacturing industries: Evidence from the Annual survey of manufacturing.” Quarterly Journal of Economics 109, 367-98. Bertrand, Marianne and Sendhil Mullainathan. (2003). “Are Emily and Greg More Employable than Lakisha and Jamal?” A Field Experiment on Labor Market Discrimination. Cambridge, Massachusetts: National Bureau of Economic Research, working paper 9873.

30

Bertola, Giuseppe (1990). “Job Security, Employment and Wages.” European Economic Review 34, 851 – 886.

Bertola, G. (1991) “Labour turnover cost and Average Labour Demand” NBER Working Paper No. 3866.

Blomstrom, M., Fors, G. and Lipsey, R. E. (1997). “Foreign Direct Investment and

Employment: Home Country experience in the United States and Sweden.” NBER Working

Paper Series 6205. www.nber.org/papers/w6205. Boug, Pål (1999). “The Demand for Labour and the Lucas Critique: Evidence from Norwegian Manufacturing Statistics” Norway Research Department, Discussion Papers No. 256. Buzás, S. and Fóti, K. (2006) “Industrial Competitiveness and Labour Market Transformation in Hungary: Macroeconomic Developments and Empirical Analysis.” Institute For World Economics Hungarian Academy of Sciences W o r k i n g P a pe r s No. 174 Card, David. (1999). “The Causal Effect of Education on Earnings.” Handbook of Labor Economics 3, 1801-1863.

Chen, T. and Ku, Y. (2003). “The Effect of Overseas Investment on Domestic Employment.” NBER Working Paper 10156. Available at http://www.nber.org/papers/w10156 Cobb, Charles W. and Paul H. Douglas (1928). “A Theory of Production.” American Economic Review 18(1), pp. 139‐165. Comola, M. and L. de Mello (2009). “The Determinants of Employment and Earnings in Indonesia: A Multinomial Selection Approach.” OECD Economics Department Working Papers, No. 690, OECD Publishing. http://dx.doi.org/10.1787/224864812153 Devarajan, S., W. Easterly, and H. Pack (1999). “Is Investment in Africa Too High or Too Low? Macro and Micro Evidence”, Journal of African Economies, vol. 10, supplement 2, pp. 81-108. Dinh, H. T., Palmade, V. T., Chandra, V. and Cossar, F. (2012). “Light Manufacturing in Africa: Targeted Policies to Enhance Private Investment and create Jobs.” African Development Forum, World Bank 67209. Driffield, N. and Taylor, K. (2000). “FDI and the labour market: A review of the evidence and policy implications.” Oxford Review of Economic Policy. Vol 16, No. 3 pp 90 -103. Dunlop, John T. (1938). “The Movement of Real and Money Wage Rates.” Economic Journal, vol. 48, pp. 413-434. Emery, J. J. (2003), “Governance, Transparency and Private Investment in Africa”, OECD Global Forum on International Investment, Johannesburg, 17-18 November 2003, pp. 1-16.

31

Estache, A., Ianchovichina, E., Bacon, R. and Salamon, I. (2013). “Infrastructure and Employment Creation in the Middle East and North Africa” International Bank for Reconstruction and Development / The World Bank Policy Research Working Paper 74918. Falk, M. and B. Koebel (2004). “The impact of office machinery, and computer capital on the demand for heterogeneous labour.” Labour Economics 11 (1), 99 -117. Flaig, G. and V. Steiner (1989). “Stability and Dynamic Properties of Labour Demand in West-German Manufacturing.” Oxford Bulletin of Economics and Statistics 51, 395-412. Fouladi, M. (2010). “The Impact of Government Expenditure on GDP, Employment and Private Investment a CGE Model Approach” Iranian Economic Review, Vol.15, No.27, Fall 2010 pp. 53 – 76. Funke, M., Maurer, W. and Strulik, H. (1999). “Capital Structure and Labour Demand: Investigations Using German Micro Data.” Oxford Bulletin of Economics and Statistics, 61, 2 (1999) pp. 199 – 215. Garrett-Peltier, H. (2010). “The Employment Impacts of Economy-Wide Investments in Renewable Energy and Energy Efficiency.” an unpublished PhD Thesis submitted to the Graduate School of the University of Massachusetts Amherst. Goos, Maarten and Manning, Alan (2007) “Lousy and Lovely Jobs: The Rising Polarization of Work in Britain.” Review of Economics and Statistics, 89(1): 118-133. Görg, H. and Hanley, A. (2005) “Labour demand effects of international outsourcing: Evidence from plant level data.” International Review of Economics and Finance, Vol. 14 (3) pp. 365-376.

Goux, D., Maurin, E. and Pauchet, M. (2001). “Fixed-term contracts and the dynamics of labour demand” European Economic Review 45 (2001) 533 – 552. Greene, William H. (2008). Econometric Analysis, 6th ed. Upper Saddle River, NJ: Prentice Hall. Guiseppe, Bertola (1991). “Labour turnover cost and Average Labour Demand” NBER Working Paper No. 3866

Habib, M. D. and Sarwar, S. (2013). “Impact of Foreign Direct Investment on Employment Level In Pakistan: A Time Series Analysis.” Journal of Law, Policy and Globalization, Vol.10, 2013 pp 46-55. www.iiste.org

Hausman, J. A. (1978). “Specification Tests in Econometrics.” Econometrica, 46(6):1251- 1271. Hamermesh, D. S. (1988). “Labour Demand and Structure of Adjustment Costs.” NBER Working Paper No. 2572.

32

Hamermesh, D. (1993). “Spatial and Temporal Aggregation in the Dynamics of Labor Demand.” In: Jan, C. – Pfann, J. A. – Ridder, G. (ed): Labor Demand and Equilibrium Wage Formation. North-Holland, Amsterdam, pp. 91–108. Henisz, Witold J. (2000). “The Institutional Environment for Economic Growth,” Economics and Politics, Wiley Blackwell, vol. 12(1), pages 1-31, 03. Henisz, Witold J. (2002). “The institutional environment for infrastructure investment,” Industrial and Corporate Change, Oxford University Press, vol. 11(2), pages 355-389. Henisz, Witold J. (2010). POLCON 2010 codebook. Manuscript, University of Pennsylvania. Henneberger, F. and Ziegler, A. (2006) “Employment Effects of Foreign Direct Investment in the Service Sector: A Systematic Approach” Nr. 109 der Reihe DISKUSSIONSPAPIERE des Forschungsinstituts für Arbeit und Arbeitsrecht an der Universität St. Gallen Herr, H. Kazandziska, M. and Mahnkopf-Praprotnik, S. (2009). “The Theoretical Debate about Minimum Wages.” Global Labour University Working Papers No. 6 February, 2009.

Hijzen, A., H. Gorg, and R. C. Hine (2005). “International Outsourcing and the Skill Structure of Labour Demand In The United Kingdom.” The Economic Journal 115 (506), 860-878. Hijzen, A. and P. Swaim (2010). “Off shoring, labour market institutions and the elasticity of labour demand.” European Economic Review 54 (8), 1016 -1034.

Hsiao, Cheng, (2003). “Analysis of Panel Data,” Cambridge University Press.

Huang, M. and Ren, P. (2013). “A Study on The Employment Effect of Chinese Investment in South Africa” Centre for Chinese Studies, Discussion Paper 5/2013 Stellenbosch University | October 2013. Ianchovichina, E., Estache, A., Renaud Foucart, R., Garsous, G. and Tito Yepes, T. (2012) “Job Creation through Infrastructure Investment in the Middle East and North Africa.” World Bank Policy Research Working Paper 6164. IFC (2011). IFC Jobs Study: Assessing Private Sector Contributions to Job Creation and Poverty Reduction. IFC, Washington, D.C. ILO (2012). EU: European Union. CIS: Commonwealth of Independent States

ILO (2012). Global Employment Trends 2012. International Labor Office, Geneva IMF (2013). “Regional Economic Outlook: Sub-Saharan Africa, Building momentum in a Multi-Speed World.” World Economic and Financial Survey. Inter-Agency Working Group (2012). “Promoting responsible investment for sustainable development and job creation”- Final report to the High-Level Development Working Group on the work of the Private Investment and Job Creation Pillar of the G20 multi-year action

33

plan on development , Mexico Summit. Pp. 1-23. Accessed from http://unctad.org/en/Pages/DIAE/G-20/Private-Investment-and-Job-Creation.aspx

Inter-Agency Working Group (IAWG) (2012) “Promoting responsible investment for sustainable development and job creation” Final report to the High-Level Development Working Group (The G20 Development Working Group) on the work of the Private Investment and Job Creation Pillar at the Mexico Summit. Pg 1-23. International Labour Organisation (ILO, 2013). “Global Employment Trends: Recovering from a second jobs dip.” International Labour Office, Geneva. Jayaraman, T. K. and Singh, B. (2007) “Foreign Direct Investment and Employment Creation in Pacific Island Countries: An empirical study of Fiji.” Asia-Pacific Research and Training Network on Trade Working Paper Series, No. 35, May 2007 Jenkinson, T.J. (1986). “Testing Neoclassical Theories of Labour Demand: An Application of Cointegration Techniques.” Oxford Bulletin of Economics and Statistics 48, 231-251.

Kaplinsky, Raphael and Morris, Mike (2009). “Chinese FDI in Sub-Saharan Africa: engaging with large dragons.” European Journal of Development Research, 21(4), pp. 551–569. http://dx.doi.org/doi:10.1057/ejdr.2009.24 Karlsson, S., Lundin, N., Sjöholm, F. and He, P. (2007). “FDI and Job Creation in China.” IFN Working Paper No. 723, 2007 Katsoulacos, Y.S. (1984). “Product innovation and employment” European Econotnic Review 26, 83-108. Kessing, S. G. (2003). “A note on the determinants of labour share movements” Economics Letters 81 (2003) 9–12.

Klette, Jakob & Førre, Svein Erik (1998). “Innovation and Job Creation In A Small open Economy-Evidence From Norwegian Manufacturing Plants 1982–92,” Economics of Innovation and New Technology, Vol. 5:2-4, pp. 247-272, DOI: 10.1080/10438599800000007 Köllo, J. – G. Korösi – É. Surányi (2003). “Labour – the Demand Side, in: Fazekas, K. and J. Koltay (szerk.)” The Hungarian Labour Market, Review and Analysis, 2003; IEHAS, 91–131. old.

Lazear, Edward P. (1990). “Job Security Provisions and Employment,” Quarterly Journal of Economics CV, 699 – 726. Lichter, A., Peichl, A. and Siegloch, S. (2013). “Labor Demand Elasticities in Europe: A Meta-Analysis.” Neujobs Working Paper No. D10.7 pp. 1- 30. Machin, S.J., A. Ryan and J. Van Reenen (1996). “Technology and changes in the skill structure: evidence from an international panel of industries.” Discussion Paper, Centre for Economic Policy Research.

34

Mairesse, J. – Dormont, B. (1985). “Labor and Investment Demand at the Firm Level: A Comparison of French, German and U.S. Manufacturing, 1970–79.” European Economic Review, June-July, Vol. 28. No. 1– 2. pp. 201–231. Malik, S., Chaudhry, S. I. and Javed, H. I. (2011). “Globalization and Employment: Evidence from Pakistan.” Pakistan Journal of Social Sciences (PJSS), Vol. 31, No. 2 (December 2011), pp. 215-226.

Manning, Alan (2013) ‘Lousy and Lovely Jobs.’ London School of Economics, Centre for Economic Performance, Paper No. CEPCP398.

Masso, J., Varblane, U. and Vahter, P. (2007). “The Impact of Outward FDI on Home-Country Employment in a Low-Cost Transition Economy.” William Davidson Institute Working Paper Number 873 Mehmet M. and Demirsel, M. T. (2013). “The Effect of Foreign Direct Investments on Unemployment: Evidence from Panel Data for Seven Developing Countries” Journal of Business, Economics and Finance Vol. 2, Issue 3, pp. 53 – 66. Michaillat, Pascal and Saez, Emmanuel (2013). “A Model of Aggregate Demand and Unemployment” London School of Economics, Centre for Economic Performance Paper No. CEPDP1235. Mileva, Eliva (2007). “Using Arellano – Bond Dynamic GMM Estimators in Stata: Tutorials with Examples using Stata 9.0” (Retrieved from the internet on 2/9/2013). Ndikumana, L. and Verick, S. (2008). “ The linkages between FDI and domestic investment: unravelling the developmental impact of foreign investment in Sub-Saharan Africa,” IZA Discussion Papers, No. 3296. http://hdl.handle.net/10419/35246 Neumark, D. and Wascher, W. (2006). “Minimum Wages and Employment: A Review of Evidence From the New Minimum Wage Research.” National Bureau of Economic Research, Working Paper 12663. http://www.nber.org/papers/w12663 Neumark, D. and Wascher, W. L.(2007). “Minimum Wages and Employment” Foundations and Trends in Microeconomics Vol. 3, No 1–2 (2007) 1–182 DOI: 10.1561/0700000015 Nickell, S. (1986). Dynamic Models of Labour Demand. In: Ashenfelter, O. – Layard, R. (ed): Handbook of Labour Economics. Volume I. Elsevier Science Publishers, Amsterdam-New York. pp. 473–522. Nickell, S. and Wadhwani, S. (1991). “Employment Determination in British Industry: Investigations Using Micro- Data.” Review of Economic Studies, October, Vol. 58. No. 5. pp. 329–345. Nickell, Stephen (2010). “The Unemployment Challenge in Europe,” CESifo Forum, Ifo Institute for Economic Research at the University of Munich, vol. 11(1), pages 3-6, 04.

35

OECD-ILO Report (2008). “Employment and Industrial Relations – Promoting responsible business conduct in a Globalising Economy.” Conference on Corporate Social Responsibility held in Paris on 23-24 June 2008.

Park, Hun Myoung, (2011). Regression Models for Panel Data Using Stata. Available at http://www.iuj.ac.jp/faculty/kucc625 Accessed on 24th November, 2013

Peichl, A. and S. Siegloch (2012). “Accounting for labor demand effects in structural labor supply models.” Labour Economics 19 (1), 129 -138. Pereira, A. M. And Andraz, J. M. (2012). “On the Economic Effects of Investment in Railroad Infrastructures in Portugal” Journal Of Economic Development, Volume 37, Number 2, pp. 79 – 107. Pierluigi, B and Roma, M. (2008). “Labour cost and employment across euro area countries and sectors” European Central Bank Working Paper Series No. 912 Pryor, Frederic & David Schaffer. (1999). “Who's Not Working and Why.” Cambridge, Massachusetts: Cambridge University Press. Psaltopoulos, D.,Skuras, D. and Thomson, K. J. (2011). “Employment effects of private investment initiatives in rural areas of southern Europe: a regional SAM approach.” Agricultural Economics Review, vol. 12 No.2 pp.50-61. Say, Jean Baptiste (1826). Traite’ d’economie politique. Philadelphia. Sackey, H. A. (2007). “Private Investment for Structural Transformation and Growth in Africa: Where do Small and Medium-Sized Enterprises Stand?” Proceedings of the African Economic Conference, pp. 371-398. Schumpeter, J.A. (1943). Capitcllism, socialism and democracy. George Allen & Unwin Ltd., U.K. Singer, M. (1996). “Dynamic Labor Demand Estimation and Stability of Coefficients – The Case of the Czech Republic.” CERGE-EI WP No. 99. Sparrow, G. N., Ortmann, G. F., Lyne, M. C. and Darroch, M. A. G. (2008). “Determinants of the demand for Regular Farm Labour in South Africa, 1960-2002.” Agrekon, Vol. 47, No. 1, pp. 52-75. Symons, J. and Layard, R. (1984). “Neoclassical Demand for Labour Functions for Six Major Economies.” The Economic Journal, vol. 94, No. 376, pp. 788-799. URL: http://www.jstor.org/stable/2232295

Symons, J. S. (1982). “Relative Prices and the Demand for Labour in British Manufacturing.” London School of Economics, Centre for Labour Economics, Discussion Paper No. 137.

Torres-Reyna,Oscar (2013). “Panel Data Analysis Fixed and Random Effects -Using Stata 10x (Ver. 4.1).” Data and Statistical Services, Princeton University. Available at http://dss.princeton.edu/training/ Accessed on 13th November, 2013.

36

Van Reenen, John and Bond, Steve (2005). “Micro-economic models of investment and employment” In: Heckman, Jim and Leamer, E, (eds.) Handbooks of econometrics volume VI. Elsevier Science Ltd, Amsterdam. ISBN 0-444506314 Van Reenen, J. (1997). “Employment and Technological Innovation: Evidence from U.K. Manufacturing Firms.” Journal of Labor Economics 15 (2), 255 - 284. Wolman, H. Levy, A., Young, G. & P. Blumenthal. (2008). “Economic Competitiveness and the Determinants of Sub-national Area Economic Activity.” Washington, DC: Prepared for the Office of Revenue Analysis in the DC Office of the Chief Financial Officer. World Bank (2004) MENA Development Report: Unlocking the Employment Potential in the Middle East and North Africa-Towards a New Social Contract. No. 28815.

World Bank (2012). World Development Report 2013: Jobs World Bank, Washington, D.C.

37

APENDIXES

Appendix 1

38

Appendix 2

Construction of the Governance Index using principal Component Analysis

GOVt = W1CCt +W2GEt + W3PSt+ W4RQt + W5RLt+ W6VAt

Where the components have been explained in the Table 1 below:

Appendix 2: Components of Country Governance Index Variable Meaning Measurement CC Control of Corruption Number of sources GE Government Effectiveness Number of Sources PS Political Stability Number of sources RQ Regulatory Quality Number of sources RL Rule of Law Number of sources VA Voice and Accountability Number of sources

The various eigenvalues, eigenvectors of the variance proportion and weights of the country governance variables have been shown in Appendix 3 and 4 below. Meanwhile the variance proportions of the various countries used in the study, as shown in Appendix 3 depict that, in all the countries, the first composition gives the best weights to be used in the calculation of the governance index.

Appendix 3: Countries and the Variance Proportions (with compositions) of the Governance index Variables:

COUNTRY VARIANCE PROPORTION (EIGENVALUES)

Comp 1 Comp 2 Comp 3 Comp 4 Comp 5 Comp 6 ALGERIA 0.975202 0.014306 0.005737 0.003317 0.000931 0.000507

39

(39.1706) (0.5746) (0.230448) (0.133216) (0.037384) (0.020384) ANGOLA 0.978235 0.010739 0.00471 0.004005 0.001924 0.000387

(34.62543) (0.380113) (0.166707) (0.141764) (0.068116) (0.013707) BENIN 0.986398 0.005011 0.004357 0.002746 0.001119 0.00037

(65.71188) (0.333805) (0.290241) (0.182954) (0.074552) 0.024619) BOTSWANA 0.967676 0.018983 0.00924 0.00329 0.000695 0.000117

(43.01453) (0.843815) (0.410715) (0.146228) (0.03091) (0.005192) BURKINA FASO 0.989244 0.005125 0.003257 0.001325 0.000791 0.000257

(68.75936) (0.356204) (0.226412) (0.092118) (0.059412) (0.017885) BURUNDI 0.985925 0.006925 0.003032 0.002635 0.001356 0.000127

(43.19584) (0.303397) (0.132842) (0.115434) (0.05941) (0.005572) CAMEROON 0.982601 0.009951 0.005214 0.001554 0.000536 0.000145

(44.51727) (0.450817) (0.236232) (0.070394) (0.024263) (0.006583) CAPE VERDE 0.947932 0.026902 0.016463 0.005191 0.003306 0.000207

(17.53282) (0.497581) (0.30449) (0.096017) (0.061139) (0.003821) CENT.AFR. REP. 0.939897 0.033598 0.019081 0.004432 0.001872 0.00112

(19.15039) (0.684553) (0.388779) (0.090301) (0.038147) (0.022827) CHAD 0.988629 0.007077 0.003263 0.000685 0.00029 0.000056

(49.5825) (0.354948) (0.16365) (0.034336) (0.014545) (0.0022827) COMOROS 0.938028 0.040454 0.014495 0.005504 0.001519 0

(10.63099) (0.458476) (0.164274) (0.062377) (0.021496) 0.00000 CONGO DR 0.958963 0.030725 0.004812 0.002555 0.002107 0.000838

(35.59486) (1.140454) (0.178608) (0.094836) (0.07819) (0.031108) CONGO REP 0.974778 0.012431 0.008732 0.002181 0.001745 0.000134

(27.36826) (0.349009) (0.245159) (0.061224) (0.04899) (0.003751) COTE D' VOIRE 0.973219 0.013162 0.00688 0.005839 0.000736 0.000163

(28.40584) (0.384174) (0.200807) (0.170426) (0.021496) (0.004756) DJIBOUTI 0.941884 0.041458 0.011354 0.00335 0.001955 0

(15.54762) (0.684338) (0.187416) (0.055302) (0.032264) 0.0000 EGYPT 0.980422 0.010436 0.006009 0.001559 0.00094 0.000635

(40.02707) (0.426066) (0.24533) (0.063629) (0.03837) (0.025922) EQUITORIAL GUINEA 0.847051 0.126633 0.020288 0.005026 0.001002 0

(6.864644) (1.026256) (0.164418) (0.040728) (0.008121) 0.00000 ERITREA 0.964407 0.015899 0.010687 0.007293 0.001649 0.000066

(22.48944) (0.370745) (0.249213) (0.17006) (0.038459) (0.001528) ETHIOPIA 0.979586 0.011495 0.007088 0.001044 0.000535 0.000253

(48.77523) (0.572336) (0.352922) (0.051976) (0.026631) (0.012574) GABON 0.945795 0.033079 0.016146 0.003897 0.001083 0

(10.76499) (0.376503) (0.18377) (0.044361) (0.012325) 0.00000 GAMBIA, THE 0.971402 0.019344 0.007633 0.00162 0 0

(17.66064) (0.351685) (0.138777) (0.029457) 0.00000 0.0000 GHANA 0.971362 0.016914 0.005918 0.004141 0.001412 0.000253

(54.84146) (0.9549333) (0.334093) (0.233815) (0.079728) (0.014302) GUINEA 0.974843 0.013702 0.006255 0.003592 0.001434 0.000174

(20.94558) (0.294405) (0.134389) (0.077186) (0.030815) (0.003738)

40

GUINEA BISSAU 0.883396 0.094926 0.015323 0.005634 0.000721 0 (6.834046) (0.734359) (0.118542) (0.043585) (0.00558) 0.0000

KENYA 0.981528 0.010763 0.004154 0.002427 0.001046 0.000082 (65.38068) (0.716944) (0.276711) (0.161663) (0.069676) (0.005433)

LESOTHO 0.955286 0.023103 0.011685 0.006681 0.003206 0.000039 (30.28391) (0.732382) (0.370425) (0.211809) (0.101638) (0.001227)

LIBERIA 0.971763 0.018321 0.00443 0.00367 0.001507 0.000309 (45.49066) (0.857652) (0.20738) (0.171795) (0.070533) (0.014481)

LIBYA 0.948801 0.034918 0.009639 0.005421 0.001222 0 (18.1524) (0.668046) (0.184404) (0.103714) (0.023381) 0.0000

MADAGASCAR 0.984902 0.006635 0.004123 0.002433 0.001328 0.000579 (60.16794) (0.405326) (0.251902) (0.148603) (0.081127) (0.035379)

MALAWI 0.959983 0.026791 0.005558 0.004745 0.002517 0.000405 (45.47254) (1.269039) (0.263288) (0.224765) (0.119219) (0.019207)

MALI 0.974483 0.011684 0.007066 0.004178 0.002074 0.000514 (51.30247) (0.615113) (0.372018) (0.219952) (0.109208) (0.027073)

MAURITANIA 0.988453 0.006923 0.002982 0.001356 0.000198 0.000088 (54.08349) (0.378797) (0.163143) (0.07417) (0.010845) (0.004838)

MAURITIUS 0.981191 0.012902 0.00478 0.00107 0.000057 0 (32.27028) (0.424341) (0.157209) (0.035191) (0.001866) (0.0000)

MOROCCO 0.980809 0.01009 0.006633 0.001818 0.000386 0.000265 (48.58408) (0.499806) (0.328549) (0.090047) (0.000386) (0.000265)

MOZAMBIQUE 0.979319 0.012371 0.004052 0.002917 0.001309 0.000032 (60.21454) (0.760619) (0.249155) (0.179349) (0.080472) (0.001977)

NAMIBIA 0.945223 0.03107 0.017122 0.004625 0.001857 0.000104 (29.70231) (0.97632) (0.538055) (0.145322) (0.058368) (0.003359)

NIGER 0.973272 0.017376 0.007181 0.00163 0.000356 0.000185 (30.46881) (0.543961) (0.224803) (0.051036) (0.011159) (0.005787)

NIGERIA 0.977411 0.01352 0.004355 0.003575 0.000689 0.000451 (59.45915) (0.822458) (0.264907) (0.217472) (0.04193) (0.027412)

RWANDA 0.971954 0.01501 0.007693 0.004835 0.000364 0.000144 (37.4) (0.577559) (0.296031) (0.186035) (0.014021) (0.005525)

SAO TOME 0.916938 0.055164 0.021199 0.005098 0.001601 0 (7.717562) (0.464293) (0.178426) (0.042909) (0.013477) (0.00000)

SENEGAL 0.974674 0.013769 0.006554 0.003715 0.000768 0.00052 (49.90464) (0.704995) (0.335592) (0.190218) (0.039317) (0.026623)

SEYCHELES 0.926622 0.035339 0.023489 0.012459 0.002091 0 (10.77197) (0.41082) (0.273057) (0.144839) (0.024309) (0.0000)

SIERRA LEONE 0.963631 0.022395 0.010374 0.002462 0.001002 0.000136 (35.5406) (0.825958) (0.382608) (0.090813) (0.036967) (0.005002)

SOMALIA 0.887752 0.07268 0.021776 0.008362 0.006807 0.002623 (9.660467) (0.790904) (0.236966) (0.090994) (0.074073) (0.028541)

SOUTH AFRICA 0.916938 0.055164 0.021199 0.005098 0.001601 0 (7.717562) (0.464293) (0.178426) (0.042909) (0.013477) (0.00000)

SUDAN 0.95912 0.022424 0.010185 0.006161 0.00114 0.00097 (19.595315) (0.458137) (0.208091) (0.125871) (0.023281) (0.019821)

41

SWAZILAND 0.954352 0.025192 0.015914 0.003698 0.000845 0 (15.11057) (0.398872) (0.251965) (0.058546) (0.013379) (0.0000)

TANZANIA 0.97153 0.018242 0.004771 0.003007 0.002185 0.000265 (54.4799) (1.022928) (0.267523) (0.168644) (0.12254) (0.014854)

TOGO 0.954674 0.026752 0.013018 0.003886 0.00114 0.000529 (24.11215) (0.675677) (0.328806) (0.098152) (0.028786) (0.013372)

TUNISIA 0.982821 0.007898 0.005225 0.002405 0.001337 0.000314 (37.04688) (0.297702) (0.196951) (0.090667) (0.050397) (0.011849)

UGANDA 0.977318 0.012995 0.005158 0.003189 0.001186 0.000153 (63.89219) (0.849566) (0.337221) (0.208478) (0.077514) (0.010034)

ZAMBIA 0.971613 0.014699 0.007855 0.003983 0.001328 0.000521 (47.66978) (0.721161) (0.385378) (0.195427) (0.065172) (0.025577)

ZIMBABWE 0.98148 0.008507 0.005367 0.003258 0.000911 0.000478 (49.70974) (0.798258) (0.426410) (0.233660) (0.075570) (0.029093)

Source: Author’s Estimation

Appendix 4: Countries and Their Estimated Governance Variables weights

COUNTRY CC GE PS RQ RL VA Algeria 0.425667 0.307707 0.203866 0.244882 0.508274 0.603534 Angola 0.456488 0.322439 0.146312 0.282654 0.515245 0.566457 Benin 0.437982 0.406387 0.222819 0.317676 0.456997 0.53255 Botswana 0.400661 0.408421 0.183787 0.307917 0.440988 0.591272 Burkina 0.437169 0.37803 0.218604 0.286286 0.456459 0.572604 Burundi 0.429352 0.362189 0.204694 0.326784 0.502805 0.531955 Cameroon 0.480414 0.351414 0.192791 0.286081 0.443692 0.574314 Cape Verde 0.358565 0.37767 0.246968 0.315205 0.47284 0.587258 CAR -0.36672 -0.35992 -0.2073 -0.36292 -0.52878 -0.53074 Chad 0.403958 0.394241 0.244565 0.332039 0.456839 0.550116 Comoros -0.3065 -0.3065 -0.07215 -0.32575 -0.55791 -0.62413 Congo DR -0.40367 -0.32693 -0.16281 -0.27863 -0.44739 -0.65258 Congo REP 0.393766 0.403557 0.217138 0.35828 0.473452 0.531432 Cote d' Voire 0.433773 0.374442 0.213661 0.322381 0.469697 0.549035 Djibouti -0.32765 -0.32765 -0.16414 -0.24644 -0.53443 -0.64188 Egypt 0.415842 0.30509 0.201733 0.238124 0.482994 0.63507 Equitorial Guinea -0.23159 -0.23159 -0.06847 -0.30931 -0.59329 -0.66361 Eritrea 0.413449 0.309297 0.177014 0.371781 0.527685 0.534218 Ethiopia -0.47486 -0.33565 -0.18415 -0.27525 -0.49649 -0.55288 Gabon -0.44305 -0.36557 -0.26022 -0.36557 -0.44584 -0.51954 Gambia -0.37981 -0.37981 -0.23596 -0.37981 -0.46377 -0.54449 Ghana -0.48618 -0.345 -0.15396 -0.24926 -0.46442 -0.58574 Guinea -0.40571 -0.395 -0.18522 -0.33199 -0.48261 -0.54949 Guinea Bissau -0.31955 -0.31955 -0.05339 -0.39791 -0.54842 -0.57779 Kenya 0.488712 0.334584 0.158106 0.252173 0.481081 0.573747 Lesotho 0.406304 0.420281 0.200793 0.36525 0.486648 0.497725 Liberia 0.48506 0.314574 0.09345 0.253119 0.530404 0.558238 Libya -0.42233 -0.31232 -0.25552 -0.31232 -0.50697 -0.55159

42

Madagascar 0.426149 0.414889 0.222283 0.322876 0.467988 0.523061 Malawi -0.44217 -0.33235 -0.11916 -0.23404 -0.49482 -0.61661 Mali 0.415859 0.403241 0.204167 0.306504 0.45491 0.56735 Mauritania 0.430902 0.371881 0.2132 0.309173 0.480543 0.55142 Mauritius 0.396084 0.396084 0.223709 0.428648 0.497039 0.453212 Morocco 0.492434 0.328635 0.173912 0.269538 0.464485 0.575209 Mozambique 0.454734 0.348465 0.164827 0.261697 0.504776 0.566866 Namibia -0.39361 -0.40611 -0.18407 -0.29981 -0.44709 -0.59707 Niger -0.39709 -0.38835 -0.19489 -0.3086 -0.44917 -0.5971 Nigeria 0.490345 0.333739 0.156702 0.247654 0.473168 0.581726 Rwanda -0.43802 -0.33285 -0.16144 -0.26271 -0.49527 -0.59748 SaoTome -0.32287 -0.32287 -0.19811 -0.4123 -0.55673 -0.52185 Senegal 0.477937 0.371043 0.170143 0.270023 0.472129 0.556 Seychelles -0.32289 -0.32289 -0.27179 -0.39114 -0.53548 -0.52715 Sierraleone -0.48069 -0.28505 -0.10668 -0.23155 -0.48442 -0.62292 Somalia -0.4934 -0.37533 -0.15626 -0.21621 -0.30452 -0.67215 South Africa -0.32287 -0.32287 -0.19811 -0.4123 -0.55673 -0.52185 Sudan -0.43399 -0.37717 -0.14367 -0.23345 -0.42355 -0.6441 Swaziland -0.44918 -0.33496 -0.19454 -0.33496 -0.52281 -0.51251 Tanzania -0.44754 -0.33544 -0.14274 -0.24268 -0.4974 -0.60042 Togo -0.43251 -0.39279 -0.21177 -0.34468 -0.48068 -0.51377 Tunisia 0.434088 0.365752 0.310332 0.198675 0.488128 0.551132 Uganda 0.472367 0.338244 0.159668 0.253691 0.473576 0.590198 Zambia -0.45617 -0.35935 -0.15175 -0.26166 -0.47544 -0.58757 Zimbabwe 0.443951 0.324388 0.183544 0.260236 0.48575 0.600263 Source: Author’s Estimation

Appendix 5: Summary Statistics, Countries that recorded lowest and highest employment and investment levels over the study period

Minimum (Country and Year) Maximum (Country and Year) Empfem Mauritania (12.70% : 1991) Rwanda (88.2% : 1991) Empmal South Africa (44.1% : 2003) Ethiopia (88.6% : 2005) Emptot Mauritania (31.8% : 1991) Rwanda (88.3%: 1991) Empfemy Mauritania (6.1%: 1991) Rwanda (81.1%: 1991 Empmal South Africa (13.9%: 2003) Baukina Faso (79.7% : 1991, 95) Emptoty Namibia (10.5% : 2009) Rwanda (80%: 1991) Total Investment Zimbabwe (2.00044% : 2005) Equit. Guinea (113.578%: 1996) GPINV Congo Dem Rep (0.100101: 1998) Equit. Guinea (42.9755% : 2009) PRINV Liberia (-2.64039: 2001) Equit. Guinea (112.352% : 1996)