Sources of Instability in Nigeria’s Manufacturing Performance: 1981

Abstract

The article investigates factors that may be responsible for observed instability in Nigeria’s manufacturing sector performance. Based on growing concerns regarding early deindustrialisation observed for many developing countries, the study examines how a menu of fiscal and monetary policies can be applied to revitalise the Nigerian manufacturing sector. Consequently, we construct an index of manufacturing sector instability and examine how it responds to a mixture of fiscal and monetary policy variables. We use annual time series data from 1981–2018 and apply the ARDL bounds test technique. Our findings show that budget deficits induce instability in the performance of the Nigerian manufacturing sector, while government infrastructural investments stabilise it. The monetary policy instruments were found to have inconsistent short-run and long-run influences but mostly conform with theory.

JEL Codes: F63, L20, L25, L60, O10, O14

Keywords

Growth catalytic performance management instability

I. Introduction

The vital role of manufacturing as an engine of growth and development is well acknowledged in the economic literature. Intuitively, the condition of a country’s manufacturing sector points to its level of technological advancement and development. Studies have shown that a vibrant manufacturing sector creates employment opportunities, increases foreign exchange earnings and reduces dependence on imported finished goods (Herman, 2016; Kaldor, 1966; Marconi et al., 2016; Thirlwall, 1983). The manufacturing sector acts as a catalyst for growth of other sectors through backward and forward linkages. The traditional view in the related literature is that the manufacturing sector is more critical in developing countries in their quest to industrialise than in developed countries due to the close relationship between manufacturing and industrialisation (Kaldor, 1966; Szirmai & Verspagen, 2015). According to this view, the inevitable structural changes in the initial stage of development in a typical developing economy begin when industrialisation and manufacturing activities are low. As the country progresses on a path of increased growth, industrial activities tend to increase due to rising manufacturing contribution to total output. While this occurs, the country witnesses a gradual decline in primary sector contribution to employment and production while that of secondary and tertiary sectors increase (Dasgupta & Singh, 2007; Rowthorn & Ramaswamy, 1997). This implies that the role of manufacturing in the economic progress of developing countries is indeed critical.

However, recent evidence shows that most developing countries have witnessed a gradual decline in the share of manufacturing output in Gross Domestic Product (GDP) at low per capita output levels (Dasgupta & Singh, 2007; Ghani & O’Connell, 2014; Palma, 2005; Rodrik, 2017). These countries have also experienced varying levels of instability and declining value-added in their manufacturing sectors compared to other sectors (Haraguchi et al., 2017). Expectedly, these occurrences raise questions about whether the manufacturing sector continues to be critical to the economic progress of developing and emerging countries. Some scholars believe that industrialisation in many developing countries may be unstable and problematic because these countries have only attained premature industrialisation in the first place (Rodrik, 2017). One of the major concerns of premature deindustrialisation in developing countries is the dependence on import for industrial capital, spare parts and consumer goods, which puts pressure on exchange rate and balance of payments. These deficiencies are critical, but they are being masked by the unprecedented development in Information and Communication Technology (ICT) and other services sectors of developing countries (Agumbayeva & Abdirov, 2019; Dasgupta & Singh, 2007). As observed by Agumbayeva and Abdirov (2019), ‘an economy that creates predominantly virtual values in the ICT sector is far from efficient and extremely vulnerable’.

Much attention has been paid to the relevance of the manufacturing sector in developing countries (Chakravarty & Mitra, 2009; Kathuria & Natarajan, 2013; Marconi et al., 2016; Necmi, 1999; Su & Yao, 2017; Szirmai & Verspagen, 2015, etc.). However, the issue of manufacturing instability and its causal factors have received relatively little attention in recent times. Given the importance of the manufacturing sector, policy effort to ensure stable manufacturing performance will undoubtedly enhance the prospect for growth and development of developing countries. In this regard, the Nigerian manufacturing sector is of interest to this study because of Nigeria’s visibility as the biggest economy in the African continent. Hence, the study investigates the sources of instability in Nigeria’s manufacturing performance over the period 1981–2018. To achieve this objective, we have developed a manufacturing instability (MANI) index for Nigeria following the framework adopted by Liew et al. (2018). Aside from serving as a helpful reference point for further study, the results of the article will assist policy formulations in stemming the tide of poor performance of the Nigerian manufacturing sector over the years.

The rest of the article is organised as follows: Section II discusses issues affecting manufacturing performance as revealed in the relevant literature; Section III examines performance of Nigeria’s manufacturing sector; Section IV deals with the method of data analysis and model estimation; Section V contains data analysis and results; and the article is concluded in Section VI.

II. Literature Review

One of the earliest empirical studies on manufacturing as a driver of growth and development is the seminal paper of Kaldor (1966) which claimed that a nation’s development trajectory is closely aligned to her manufacturing and industrial activities. Typically, development begins with a highly positive relationship between economic growth and manufacturing sector output. The faster the manufacturing sector grows, the faster the output growth. Higher output growth in the industry, in turn, catalyses higher productivity and employment growth in other sectors. Kaldor explains that the positive relationship between manufacturing and aggregate output tapers off beyond a certain point as diminishing returns set in. The overall implication is that countries in the early stages of development should experience a rising contribution of manufacturing output to GDP, while countries with high per capita GDP levels should experience a declining contribution of manufacturing to GDP.

Various attempts have been made over the years to validate Kaldor’s propositions with mostly consistent support documented in the evidence on developing countries (e.g., Felipe (1984) for 5 AESAN countries; Fagerberg and Verspagen (1999) for selected developing countries; Necmi (1999) for 45 developing countries; Wells and Thirlwall (2003) for selected African countries; Dasgupta and Singh (2007) for 48 developing economies; Chakravarty and Mitra (2009) and Kathuria and Natarajan (2013) for the Indian economy; Kathuria and Natarajan (2013) across 15 states in India; Rodrik (2016), for developing Asian countries; Szirmai and Verspagen (2015) for low- and middle-income countries; Su and Yao (2017) for some middle-income countries; Keho (2018) covering Economic Community of West African States [ECOWAS]; among others). There are also supporting empirical evidence concerning the developed countries (e.g., Rowthorn and Ramaswamy (1997) for 21 industrialised economies; Pons-Novell and Viladecans-Marsal (1999) for European countries; McCausland and Theodossiou (2012) for a selection of developed countries; Güçlü (2013) and Pata and Zengin (2020) covering Turkey).

In recent times, declining manufacturing value-added and its proportion to total output have been observed for developing countries, thus contradicting the established theoretical pattern. This phenomenon, known as early or premature deindustrialisation, has attracted much interest in its causes and consequences. Foellmi and Zweimüller (2011) argued that the dualistic structure of developing economies and the associated inequalities are possible triggers of premature industrialisation. This proposition has received much support in the empirical literature, with some studies pointing to decreased manufacturing performance accompanied by increasing inequality in the particular countries (e.g., Moustafa, 2006 for Egypt; Rydzck, 2013 for 65 developing countries). Other studies also indicate that manufacturing performance is susceptible to macroeconomic shocks (Ndikumana, 2008; Tkalec & Vizek, 2009; Varela et al., 2012). In particular, Varela et al. (2012) found that the Indonesian manufacturing sector’s stagnation was partially due to the challenging macroeconomic uncertainty which negatively impacted profits. This finding was corroborated by Idoko and Taiga (2018) who found that FDI in challenging macroeconomic environment has significant positive effect on Nigeria’s manufacturing performance. Also, Quoreshi and Stone (2019) found that Swedish manufacturing performance was affected by macroeconomic shocks.

From the fiscal perspective, there appears to be strong linkage between government capital expenditure and manufacturing performance. Increased government capital expenditure enhances infrastructural development, a critical component of the manufacturing sector’s growth and a development enabler (Elhiraika et al., 2014; Tybout, 2000). The ripple effect manifests in lower production cost, reduced overhead costs, cheaper raw materials cost and other handling costs. Examining the impact of macroeconomic policies on manufacturing production in Croatia, Tkalec and Vizek (2009) found that fiscal policy, more than monetary policy, was of particular importance to manufacturing performance, and crowding-out effect of increased government spending on manufacturing performance. Also, Hammed and Arawomo (2020) found a significant positive short-run response of Nigeria’s manufacturing performance to fiscal shocks. However, Ndikumana (2008) documented a ‘crowding-in’ effect of increased government spending on Nigeria’s manufacturing firms. Other Nigerian studies found that government capital investments increased manufacturing performance (Emmanuel & Oladiran, 2015; Mesagan & Ezeji, 2016). Oladipo et al. (2019) documented mixed evidence for different tax types, with corporate tax inducing a positive change and value-added tax causing an adverse change in Nigeria’s manufacturing performance.

From a monetary perspective, variables influencing manufacturing sector growth have been reported to include interest rate, exchange rate and inflation (Aregbeyen, 2012). The monetary variables play significant roles in manufacturing growth because monetary policy instruments determine the cost of capital, availability of funds, the structure of capital holdings in the manufacturing sector and so on. In this regard, Ono (2001) emphasised the important role of monetary policy in manufacturing performance. Moreover, fluctuations in the value of monetary instruments over the business cycle could result in uncertainty, which constitutes profound discouragement to the risk-averse investors interested in the sector. This argument was supported by Ibrahim (2005), who reported a substantial response to interest rate shocks for the Malaysian manufacturing sector. Asaleye et al. (2018) found that prime interest rate and private sector credit substantially influenced manufacturing sector contribution to employment and output. In a panel study on African countries, Akinyemi et al. (2018) reported that manufacturing performance increased with increased liquidity ratio and money supply but decreased with the lending rate. Hammed and Arawomo (2020) equally showed that Nigeria’s manufacturing performance responded adversely to domestic price shocks. Conversely, Asaleye et al. (2021) found no evidence that exchange rate influenced Nigeria’s manufacturing performance.

III. Performance of the Nigerian Manufacturing Sector

A review of key indices of Nigeria’s manufacturing performance over the study period shows high volatility. Manufacturing value added (MVA) fell sharply from US$38.85 billion in 1981 to US$20.88 billion in 1986, a space of six years. There was another sharp fall from US$31.06 billion in 1992 to US$19.28 billion in 1995 within a space of three years. A sharp rise was witnessed from 1999 up to 2003, followed by an erratic movement of decline and growth towards 2005 before rising sluggishly but steadily to peak at US$42.72 billion in 2018. As indicated in Figure 1, there was marked instability in manufacturing contribution to GDP (MVGP) between 1981 and 1998. The 18 years of high volatility was followed by a steady decline from 17.45 per cent in 1998 to 6.55 per cent in 2010, then rising to 9.65 per cent in 2018.

Figure 1.

Source: Authors’ computations.

Figure 1 also shows a continuous decline in manufacturing capacity utilisation rate (MCU), falling sharply from 73.3 per cent in 1981 to 29.3 per cent in 1995 and rising sharply to 56.5 per cent in 2003. From 2004 up to 2018, MCU fluctuated between 55.7 per cent and 55.1 per cent. Also, the annual growth rate of manufacturing value added (MVGT) indicates high volatility of manufacturing performance over the period. As shown in Figure 1, the annual growth rate fluctuated agitatedly between the lower band of –29.03 per cent and the upper band 17.31 per cent during the coverage period. Equally, manufacturing exports as a percentage of goods export shown in Figure 2 depicts high volatility between 2003 and 2016. While the period of 2006–2008 witnessed a sharp rise in manufacturing exports from 1.34 per cent to 5.59 per cent, 2014 to 2016 period witnessed a sharp drop in manufacturing exports as a percentage of goods export. In between these threshold periods are other erratic movements, falling from 8.17 per cent in 2008 to 6.55 per cent in 2010. It rose thereafter to 9.65 per cent in 2018.

Figure 2.

Source: Authors’ computations.

The reasons for the observed instability in manufacturing performance are varied, ranging from a multiplicity of government policies, macroeconomic shocks, poor infrastructural facilities and insecurity. Various industrial policies implemented over the years include Import Substitution Industrialisation (the 1960s–1970s), Indigenisation Policy (1972–1977), Structural Adjustment Programme (SAP, 1986), Trade and Finance Liberalisation Policy (1989), Small and Medium Industries Equity Investment Scheme (SMIEIS, 2000/2001), National Integrated Industrial Development (NIID, 2007), Industrial Park Development Strategy (IPDS, 2009) and Nigeria Industrial Revolution Plan (NIRP, 2014–2017). Unfortunately, these policies have been poorly implemented, abandoned or replaced with other policies by succeeding governments. While the import substitution industrialisation policy led to the robust growth of industrial production, the indigenisation policy resulted in industrial decline, accentuated by the negative effects of declining oil windfall recorded in the mid-1970s, which actually ceased to flow during this period. The foreign reserve of the country was severely depleted and many industries that relied on imported inputs had to either shut down or operate below full capacity. The mild and negligible gains in manufacturing performance recorded under structural adjustment programme was destroyed under the succeeding military government that faced international sanctions (Ekpo, 2014). Also, trade and finance liberalisation policy failed to ensure the desired market determination of exchange rate and nurturing of industrial productivity and efficiency through deregulated interest rates.

The SMIEIS set up to develop and package viable industries has been plagued with high cost of pre-investment activities such as feasibility studies, assets valuation, management fees, etc. (which prospective entrepreneurs would want to shy away from as wasted funds), the reluctance of banks to finance long term projects, and poor state of physical infrastructures, among others. Also, the national integrated industrial development policy did not achieve much success because of bureaucratic bottlenecks in its implementation (Ekpo, 2014). The designated industrial parks lacked operational facilities such as adequate power supply, good transport network, adequate water supply for human and industrial uses, good sewage system, etc. The slow pace of work at various national integrated power project sites (not unconnected to the slow-paced disbursement of loans meant for small and medium scale enterprises by banks) led to many projects being abandoned, prompting trespasses by local residents.

With respect to macroeconomic shocks, the monocultural nature of the Nigerian economy, which relies heavily on oil revenues, has adverse effect on manufacturing performance. Equally, it can be argued that manufacturing performance was affected by the major financial and public sector reforms of 2004–2011, the global economic crisis of 2007–2011 (which precipitated the Nigerian capital market crisis from 2008), the rebasing of Nigeria’s GDP, which saw the economy as the largest in Africa as at 2013, and the successful transition of political power from a sitting president in 2015. These occurrences reflect various macroeconomic variables with implications for manufacturing instability.

Other factors that can account for instability in manufacturing performance include excessive debt finance in the capital structures of small and medium-size manufacturing, often leading to bankruptcies and foreclosures; poor infrastructural facilities in the operating environment; and unreliable power supply that forces businesses to rely on generators thereby increasing operating costs and reducing overall competitiveness and profitability. Additionally, the incidence of insecurity of life and property has become increasingly worrisome since 2008, snowballing into full-blown terrorism in the northern parts of the country since 2014. This forced many manufacturing businesses to either close down in various parts of north-central and north-eastern Nigeria or relocate to neighbouring countries such as Ghana and Benin Republic.

Given the foregoing challenges, it is not surprising that Nigeria’s manufacturing performance falls below international benchmarks. As shown in Table 1, Nigeria’s Manufacturing Value Added growth rate of 2.1 per cent in 2018 ranks 20th among 26 countries in sub-Saharan Africa. Whereas Nigeria is the second largest economy in Africa in terms of GDP, her performance as 26th on the table is far below that of South Africa, which ranks 4th on the table but considered as the third largest economy in Africa. The 2.1 per cent average growth rate of Nigeria’s manufacturing value added in 2018 also compares very poorly to the regional average of 11 per cent for Caribbean small states and other countries in Sub-Saharan Africa, and 19 per cent for low- and middle-income countries.

Table 1.

Growth in Manufacturing Value Added by Selected Country (Percentage Value).

Country	Value	Rank	Country	Value	Rank
Marshall Islands	61.6	1	Kenya	4.2	14
Senegal	19	2	Ghana	4.1	15
Cape Verde	14.4	3	Botswana	3.6	16
South Africa	12	4	Cameroon	3.6	17
Gabon	11.7	5	Guinea-Bissau	2.6	18
Guinea	9.2	6	Cote d’Ivoire	2.5	19
Uganda	8	7	Nigeria	2.1	20
Zambia	8	8	Rwanda	2	21
Zimbabwe	8	9	Niger	1.9	22
Sao Tome & Principe	7	10	Burkina Faso	1.8	23
Togo	7	11	Mozambique	1.7	24
Seychelles	6	12	Burundi	1.2	25
Madagascar	4.9	13	Gambia, The	0.3	26
Regional average values
Caribbean small states		11
Sub-Saharan Africa		11
Low & middle income		19

Source: World Bank national accounts data, and OECD National Accounts data files.

IV. Methodology

Unit Root Tests

Due to the inherent non-stationarity typical of most time series data and the consequent risk of spurious regression, we conducted unit root test on the variables used for analysis to determine their stationarity status. For robustness, we used the well-known Augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) tests. Both tests hypothesise that the relevant series is generated by an autoregressive (AR) process, where the AR coefficient is one. The generic form of the ADF test is as follows:

Δ Y_{t} = β_{0} + β_{1} t + γ y_{t - 1} + \sum_{i = 1}^{p} α_{i} Δ Y_{t - i} + u_{t},

(1)

where $γ$ is the AR coefficient hypothesised to be equal to 0, $β_{0}$ is a constant term and t is a deterministic trend. Dickey and Fuller (1981) acknowledged the likelihood of the stochastic term to be serially correlated. To resolve the problem, they devised an ADF test by including the lagged terms of the dependent variable in the relationship. Similar specification to the ADF is the test by Phillips and Perron (1988) devised to account for possible serial correlation by adjusting the t-statistic of the AR coefficient under mild assumptions (Asteriou & Hall, 2007). PP test assumes the general form:

Δ Y_{t - 1} = φ_{0} + φ_{1} t + γ y_{t - 1} + ω_{t} .

(2)

Under null hypothesis of unit root, the t-statistics of the AR coefficients are compared to critical values provided by McKinnon (1991). Rejection of the hypothesis indicates stationarity of the relevant variable, while acceptance indicates non-stationarity.

Model Specification and Estimation Strategy

The composite MANI index can be computed from the following relationship (Liew et al., 2018):

\begin{array}{l} M A N I, = \frac{1}{n} \sum_{i = 1}^{n} (\frac{Y_{i t} - min Y_{i}}{max Y_{i} - min Y_{i}}) \\ 0 \leq MANI \leq 1, \end{array}

(3)

where $Y_{i t}$ are actual values of the manufacturing performance variables i, at time t; and $m i n Y_{i}$ and $m a x Y_{i}$ , respectively represent minimum and maximum values of the series ‘Y _i ’ at time t. The MANI index value ranges from zero to one, with zero indicating manufacturing stability and one indicating extreme manufacturing instability.

The manufacturing performance variables used to compute the instability index include manufacturing value added at constant 2010 US$, manufacturing value-added annual percentage growth rate, manufacturing value-added as a percentage of GDP, manufactured goods exports as a percentage of total goods exports and average MCUs.

Explanatory variables in the model include exchange rate, interest rate, ratio of budget deficit to GDP, inflation rate and government investment in infrastructure. It is assumed that a strong currency decreases competitiveness of manufactured exports. Rising interest rate increases funding costs, while rising budget deficit adversely impacts manufacturing activities through the crowding-out effect. Also, rising inflation reduces real incomes and effective demand for manufactured goods, while rising government investment in infrastructure reduces the cost of doing business, thereby enhancing manufacturing performance.

Thus, the manufacturing instability model can be specified as follows:

{MANI}_{t} = f ({EXR}_{t}, {INT}_{t}, {BDP}_{t}, {INF}_{t}, {GIF}_{t}),

(4)

where MANI is manufacturing instability; EXR is exchange rate as a proxy for manufacturing competitiveness; INT is interest rate, as a proxy for the cost of funding; BDP is budget deficit to GDP ratio as a proxy for access to credit; INF is inflation rate as a proxy for effective demand for locally manufactured products; GIF is government investment in infrastructure, as a proxy for the cost of doing business; and t is time subscript.

Equation (4) can be estimated as a linear model using the OLS technique. However, this approach is not desirable for some reasons. First, direct application of OLS is only possible provided the unit root test indicates stationary variables at levels. Even if this is the case, such a strategy produces long-run estimates and ignores short-run dynamics between manufacturing instability and its determinants. Second, where all the variables are stationary of the same order, directly applying OLS requires differencing, which makes it impossible to estimate the long-run relationship. Another challenge arises if the variables happen to be of varying orders of integration, in which case applying simple OLS increases the chances of performing spurious regression (Asteriou & Hall, 2007).

An alternative is to estimate a Dynamic OLS (DOLS) or Fully Modified OLS (FMOLS) model. The FMOLS requires all variables to be integrated in the same order (Othman & Masih, 2015). For this reason, the DOLS is often preferred over FMOLS. However, as Panopoulou and Pittis (2004) noted, the DOLS approach fails to correct second-order asymptotic bias of cointegrating relationships. Forest and Turner (2013) have also demonstrated that DOLS estimations have higher mean square error and inferior autocorrelation properties than other similar estimators. Moreover, Monte Carlo simulations and application to actual data have shown that selecting appropriate lag length for DOLS using the usual lag selection criteria performs far less than other techniques (Forest & Turner, 2013; Panopoulou & Pittis, 2004).

A superior alternative to OLS and DOLS is the Autoregressive Distributed Lag (ARDL) bounds testing approach developed by Pesaran et al. (2001). In addition to using it to correct possible endogeneity, the ARDL approach can be applied regardless of whether the underlying model combines I(0) and I(1) variables. More importantly, the model incorporates an error correction model (ECM) derived through the simple mathematical transformation of the original ARDL model (Dantama et al., 2012). Thus, apart from incorporating long-run estimates with short-run dynamics, ARDL also provides a measure of adjustment speed towards long-run equilibrium. Additionally, the ARDL bounds test of cointegration is superior to the approach developed by Johansen and Juselius (1990) because it circumvents the complications in applying the Johansen approach to I(0) series.

These unique characteristics render the ARDL model preferable in this instance. The particular representation of the ARDL model for this study is expressed as follows:

MAN I_{t} = α + β T + \sum_{i = 1}^{p} γ_{i} MAN I_{t - i} + \sum_{i = 0}^{q_{1}} λ_{1 i} E X R_{t - i} + \sum_{i = 0}^{q_{2}} λ_{2 i} IN T_{t - i} + \sum_{i = 0}^{q_{3}} λ_{3 i} B D P_{t - i} + \sum_{i = 0}^{q_{4}} λ_{4 i} I N F_{t - i} + \sum_{i = 0}^{q_{5}} λ_{5 i} G I F_{t - i} + ε_{t},

(5)

where T represents the deterministic trend and $ε$ is a stochastic term assumed to satisfy the usual assumptions of BLUE. The above is the over-parameterised version of the ARDL model and can be re-parameterised to yield the following error correction format:

Δ MAN I_{t} = \sum_{i = 1}^{p} δ_{i} Δ M A N I_{t - i} + \sum_{i = 0}^{q_{1}} θ_{1 i} Δ E X R_{t - i} + \sum_{i = 0}^{q_{2}} θ_{2 i} Δ I N T_{t - i} + \sum_{i = 0}^{q_{3}} θ_{3 i} Δ B D P_{t - i} + \sum_{i = 0}^{q_{4}} θ_{4 i} Δ I N F_{t - i} + \sum_{i = 0}^{q_{5}} θ_{5 i} Δ G I F_{t - i} + τ E C T_{t - 1} + ε_{t},

(6)

where

τ E C T_{t - 1} = τ (M A N I_{t - 1} - α^{*} - β^{*} T - ψ_{1} E X R_{t - 1} - ψ_{2} I N T_{t - 1} - ψ_{3} B D P_{t - 1} - ψ_{4} I N F_{t - 1} - ψ_{5} G I F_{t - 1}),

${ECT}_{t - 1}$ represents the error correction term (ECT) while $ψ_{j}$ and $θ_{j i}$ are long-run and short-run coefficients, respectively, and $α^{*} = - α / τ; β^{*} = - β / τ; ψ_{j} = - λ_{j 1} / τ, j = 1, 2, \dots, 5$ .

For convergence to occur, it is required that the coefficient of the ECT ( $τ$ ) must be negative and less than 1 in absolute terms (Kripfganz & Schneider, 2018). To test for cointegration, Pesaran et al. (2001) propose computing an F-statistic using the Wald test of coefficient restrictions based on the following hypothesis:

H_{0} : τ = λ_{11} = λ_{21} = \dots = λ_{51} = 0.

The above is known as the bound test. Its decision rule is to compare the computed F-statistic to the critical I(0) and I(1) bounds suggested by Pesaran et al. (2001). The null hypothesis is rejected if the computed F-statistic exceeds the critical I(1) bound at given level of significance and degrees of freedom. When the computed F-statistic falls below the lower bound, the conclusion is that the variables are not cointegrated. The test is inconclusive when the computed F-statistic lies between the I(0) and I(1) bounds.

Data and Sources

We used annual time series data from 1981–2018 obtained from World Development Indicators (WDI) and Central Bank of Nigeria (CBN) Statistical Bulletin, 2019 edition. MANI was computed in ratios. Budget deficit to GDP (BDP) ratios were transformed to percentages to scale up because some of the ratios obtained were very small. All other variables were expressed in natural logs. To compute BDP, we divided the budget balance of each year by the corresponding GDP value. We obtained government infrastructural investment (GIF) using the natural log of government capital expenditures. A summary description of the variables used is presented in Table 2.

Table 2.

Description of Variables.

Variable	Symbol	Measurement	A Priori Signs	Source
Manufacturing instability	MANI	Ratio		Computed
Exchange rate	EXR	Natural logs		WDI
Interest rate	INT	Natural logs	–	WDI
Budget deficit	BDP	Percentages	+	CBN
Inflation rate	INF	Natural logs	±	WDI
Government infrastructural investment	GIF	Natural logs	+	CBN

Source: Authors’ computations.

V. Empirical Results and Discussion

Preliminary Analysis

The summary statistics of the variables in the empirical model are presented in Table 3. From the table, we observe that maximum and minimum values of the index are within the theoretical limits of between 0 and 1. This is also aptly displayed in Figure 3, which shows a highly volatile trend in Nigeria’s manufacturing performance during 1981–2018.

Table 3.

Summary Descriptive Statistics.

Statistic	MANI	EXR	INT	BDP	INF	GIF
Mean	0.591	4.793	2.837	–0.622	2.686	4.922
Median	0.603	4.610	2.866	–0.350	2.543	5.665
Max.	1.000	6.294	3.455	0.147	4.288	7.428
Min.	0.000	3.915	2.188	–4.223	1.684	1.411
Std. Dev.	0.196	0.620	0.295	1.077	0.701	1.992
Skewness	–0.885	0.998	–0.694	–2.747	0.836	–0.563
Kurtosis	5.085	2.944	3.224	9.249	2.781	1.781
Jarque–Bera	11.849	6.319	3.133	109.622	4.498	4.361
Prob. JB	0.003	0.042	0.209	0.000	0.105	0.113
Observations	38	38	38	38	38	38

Source: Authors’ computations.

In Table 3, we see that the p-values of INT, INF and GIF are well above the usual levels suggesting normality in distribution. Indeed, the kurtosis of INT and INF are close to three suggesting mesokurtic distributions and lending further credence to normality in distribution. On the other hand, for MANI, EXR and BDP, the null hypothesis of normality is rejected given that their JB statistics have p-values below 5 per cent. The Kurtosis of MANI and BDP are also relatively high, suggesting leptokurtic distributions with sharper central peaks relative to normal distribution.

Figure 3.

Source: Authors’ computations.

Results of unit root test to ascertain stationarity are presented in Table 4. The ADF and PP tests were conducted twice on each variable—the first time with only a constant in the test setup and the second time with a deterministic trend. The results show that the ADF test supports stationarity at levels for MANI. On the other hand, EXR only became stationary at levels when a deterministic trend was included in the ADF test setup. Both versions of the PP test failed to support stationarity for EXR at levels. The ADF and PP tests only supported stationarity for INT and GIF in their first differences, implying that these variables are strictly I(1) series. In the case of BDP, both versions of the ADF and PP tests supported stationarity at levels, although the PP version with the trend was only marginally significant. For INF, both versions of the ADF test supported stationarity at levels, but the PP test only corroborated this without a deterministic trend in the test setup.

Table 4.

Summary of Unit Root Tests.

Series	TestVersion	Augmented Dickey–Fuller		Phillips–Perron		Status
Series	TestVersion	Y	ΔY	Y	ΔY	Status
MANI	No trend	–2.892 (0.056)	–5.370 (0.000)	–2.254 (0.192)	–5.349 (0.000)	I(0)
	Trend	–2.734 (0.230)	–5.384 (0.000)	–2.170 (0.491)	–5.359 (0.000)	I(1)
EXR	No trend	–2.171 (0.219)	–4.592 (0.001)	–2.356 (0.161)	–4.455 (0.000)	I(1)
	Trend	–3.471 (0.061)	–4.596 (0.000)	–2.275 (0.436)	–4.425 (0.000)	I(0)
INT	No trend	–2.438 (0.139)	–5.956 (0.000)	–2.449 (0.136)	–5.961 (0.000)	I(1)
	Trend	–2.040 (0.561)	–6.163 (0.000)	–2.038 (0.562)	–6.165 (0.000)	I(1)
BDP	No trend	–4.083 (0.003)	–4.911 (0.000)	–5.138 (0.000)	–4.894 (0.000)	I(0)
	Trend	–3.279 (0.085)	–5.408 (0.000)	–5.439 (0.000)	–5.392 (0.000)	I(0)
INF	No trend	–3.353 (0.019)	–6.776 (0.000)	–3.234 (0.026)	–9.319 (0.000)	I(0)
	Trend	–4.393 (0.006)	–6.673 (0.000)	–3.173 (0.105)	–9.773 (0.000)	I(0)
GIF	No trend	–0.887 (0.781)	–6.272 (0.000)	–0.884 (0.782)	–6.268 (0.000)	I(1)
	Trend	–1.336 (0.863)	–6.259 (0.000)	–1.468 (0.823)	–6.262 (0.000)	I(1)

Source: Authors’ computations.

Note: p-values of adjusted t-Statistics in parentheses; Y & ΔY denote levels and first difference of the relevant variable, respectively.

Main Results

The report in Table 5 presents the results of the bounds test of cointegration. Although the test was performed post-estimation, we have reported it ahead of the estimation result to justify or refute cointegration. The test was computed using 5 degrees of freedom, representing the number of regressors in the model. When compared to the critical I(0) and I(1) bounds, the computed F-statistic of the test exceeded the upper I(1) bound, even at 1 per cent. The result thus, provides compelling evidence that there is significant long-run relationship between manufacturing instability and the regressors. Using the Akaike information criteria to estimate our ARDL model, we set the maximum lag to 3 and obtained an ARDL (2,3,2,1,1,1) estimated model.

Table 5.

Bounds Test of Cointegration.

Significance	I(0) Bound	I(1) Bound
10%	2.75	3.79
5%	3.12	4.25
2.50%	3.49	4.67
1%	3.93	5.23
Test statistic	Value	DF
F-statistic	6.44	5

Source: Authors’ computations.

The long-run estimates are presented in Table 6. The coefficient of exchange rate indicates that a percentage increase in exchange rate causes manufacturing instability to decline by 0.0035 points. Since exchange rate mimics manufacturing competitiveness in the model, this provides evidence that more stability is infused into manufacturing performance with greater competitiveness. Interest rate is not significant and does not carry the expected sign, indicating that it is not an important policy variable. The coefficient of budget deficit is significant and correctly signed, indicating that MANI index increases by around 0.57 units for a percentage point increase in fiscal deficit. The result supports the argument that rising budget deficits constitute funding constraints to the Nigerian manufacturing sector and instigate greater sector instability. The impact of inflation on manufacturing instability is negative but only marginally significant, given a 7 per cent p-value. The coefficient indicates that manufacturing instability decreases by 0.002 units for a percentage increase in inflation. The impact of GIF is negative as expected, although insignificant. A percentage increase in GIF is associated with 0.001-point decrease in the manufacturing instability.

Table 6.

Estimated Long-Run Model.

Variable	Coefficient	Std. Error	t-Statistic	Prob.
EXR	–0.354	0.178	–1.984	0.063
INT	–0.479	0.421	–1.138	0.270
BDP	0.571	0.224	2.550	0.020
INF	–0.210	0.109	–1.928	0.070
GIF	–0.091	0.087	–1.044	0.311
@TREND	0.009	0.016	0.526	0.605
Intercept	4.720	2.054	2.298	0.034

Source: Authors’ computations.

The short-run model is summarised in Table 7. It shows a highly significant ECT. The ECT coefficient is correctly signed and within the a priori bounds (i.e., it is less than 1 in absolute terms). This suggests that up to 67.9 per cent of disequilibrium in the estimated model is adjusted each year. All the variables significantly influence manufacturing instability, except inflation. A percentage increase in exchange rate is associated with a 0.0036-points rise in the manufacturing instability. This contradicts the long-run result of the model.

Another variable whose short-run impact contradicts the long-run result is interest rate. A percentage rise in interest rate leads to increased manufacturing instability by approximately 0.006 points. This evidence supports the theoretical explanation that rising cost of fund acts as constraint to manufacturing production and triggers greater degree of instability. Several other studies have found similar evidence that rising interest rate adversely affects manufacturing performance (e.g., Ibrahim, 2005 on Malaysia; Akinyemi et al., 2018, Asaleye et al., 2018; Hammed & Arawomo, 2020 on Nigeria).

Table 7.

Estimated Short-Run Model.

Variable	Coefficient	Std. Error	t-Statistic	Prob.
D(MANI(-1))	0.466	0.267	1.742	0.099
D(EXR)	0.056	0.091	0.622	0.542
D(EXR(-1))	0.359	0.122	2.945	0.009
D(EXR(-2))	0.107	0.083	1.295	0.212
D(INT)	0.148	0.234	0.634	0.534
D(INT(-1))	0.586	0.200	2.922	0.009
D(BDP)	0.540	0.159	3.392	0.003
D(INF)	0.033	0.047	0.697	0.495
D(GIF)	–0.208	0.101	–2.048	0.055
D(@TREND)	0.006	0.012	0.487	0.632
ECT(-1)	–0.679	0.215	–3.152	0.006

Source: Authors’ computations.

Concerning budget deficit, the short-run result harmonises with the long-run estimate. A percentage increase in deficit leads to 0.005 points increase in the manufacturing instability. The short-run impact of GIF is also the same in the long-run. A percentage increase in GIF leads to fall in manufacturing instability by 0.002 points. This result fits the theoretical position that more GIF facilitates conducive environment for manufacturing industries to thrive. Similar findings were previously documented for Nigeria by Emmanuel and Oladiran (2015) and Mesagan and Ezeji (2016). Also, Brox and Fader (2005) reported similar evidence for Canada.

Table 8.

Summary of Post-Estimation Diagnostics.

Test	Purpose	Statistic	p-Value
BG LM test	Serial correlation	2.2912	.3180
BPG test	Heteroskedasticity	18.2202	.3112
RESET test	Specification error	0.0004	.9834
Jarque–Bera	Normality	1.5293	.4655

Source: Authors’ computations.

Table 8 summarises the results of post-estimation diagnoses. The Breusch–Godfrey LM serial correlation test accepts the hypothesis of non-autocorrelation given that the p-values are well above the expected levels. Also, the Breusch-Pagan-Godfrey (BPG) test to detect heteroskedastic residuals accepts the hypothesis of homoscedastic residuals.

The Ramsey RESET test and the Jarque–Bera (JB) tests were used to test model misspecification and normality, respectively. The RESET test shows that the linear ARDL model adequately expresses the relationship between manufacturing instability and its determinants. The JB test supports normality in the residuals of the regression. Lastly, the CUSUM test was done to determine stability of the estimated model over time. The CUSUM plots reported in Figure 4 confirm that the estimated coefficients do not change systematically over time.

Figure 4.

Source: Authors’ computations.

VI. Concluding Remarks and Recommendation

The incidence of premature deindustrialisation in many developing countries has received much attention in recent times. Concerns over premature deindustrialisation arise because industrialisation is considered necessary and inevitable for all countries in their path to development. As such, countries that fail to develop their industrial sectors adequately before experiencing dominance of the service sector might not attain meaningful development over time. Regarding Nigeria, her manufacturing sector remains underdeveloped, with some observed instability in its performance. On the basis of these concerns, this article investigated the sources of instability in Nigeria’s manufacturing sector performance with emphasis on fiscal and monetary policy variables. We constructed an index of manufacturing instability using an array of manufacturing performance variables and observed regular fluctuations in the index over the study period, suggesting that Nigeria’s manufacturing performance has indeed been unstable. We further analysed the roles of selected fiscal and monetary variables in the observed manufacturing instability. The results support the theoretical roles of fiscal and monetary policies in the management of manufacturing performance.

Specifically, the findings indicate that the selected fiscal policy variables consistently influenced manufacturing instability along theoretical lines in the short-run and long-run. Budget deficits infused greater instability while GIFs reduced instability. On the other hand, the monetary policy variables had inconsistent short-run and long-run impact on manufacturing instability. Exchange rate had a significant but varying impact, increasing manufacturing instability in the short-run but decreasing it in the long-run. Likewise, interest rate significantly increased manufacturing instability in the short-run but had no significant long-run influence. Furthermore, inflation had positive but insignificant short-run impact, and negative but marginally significant long-run impact.

The implication of the findings for policy is that an appropriate complement of short- and long-run policy measures should be put in place to address Nigeria’s manufacturing instability. Short-run measures recommended consist of reducing real exchange and interest rates and increasing the aggregate price level through a mix of reflationary monetary and fiscal policies. Given that most manufacturing equipment and spare parts are imported, a reduced real exchange rate implies a reduction in manufacturing production costs and increased revenues. Also, reduced interest rate lowers capital cost to manufacturing firms and increases demand for manufactured goods. Another short-run policy measure that can be gleaned from the study is that a reduction in government fiscal deficits will limit the crowding-out effect of government borrowing on the manufacturing sector. Also, increased government infrastructural expenditure has a short-run positive impact on manufacturing instability, potentially through its multiplier effect on smoothening manufacturing transportation and logistics, manufacturing production and revenue generation.

On the other hand, long-run policy measures consist of increasing real exchange and interest rates while curtailing government fiscal deficits and embarking on other deflationary efforts to curb inflation. It should be noted that the policy response on government fiscal deficit is the same both in the short- and long-run. In other words, government budgetary deficits must be managed within acceptable limits over the short and long term to curtail manufacturing instability. The policy to increase real exchange rate should encourage manufacturing exports because manufacturers would earn more foreign exchange revenue for the same amount of exported goods. Increased interest rate implies increased debt service burden of government. This limits excessive deficit financing and crowding-out effect of government borrowing on manufacturing operations.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Patrick Onosode Onodje

References

Agumbayeva

, & Abdirov

(2019). The role of digitalization in the development of new industrialization of Kazakhstan [Paper presentation]. 1st International Scientific Conference ‘Modern Management Trends and the Digital Economy: From Regional Development to Global Economic Growth’ (MTDE 2019) (pp. 326–330). Atlantis Press.

Akinyemi

A. K.

, Ogbuji

A. I.

, & Adedokun

A. S.

(2018). Manufacturing sector performance in Africa: The role of monetary policy instruments. BizEcons Quarterly , 3, 27–35. https://econpapers.repec.org/article/risbuecqu/0006.htm

Aregbeyen

(2012). The determinants of firm growth in Nigeria. Pakistan Journal of Applied Economics , 22(1–2), 19–38.

Asaleye

A. J.

, Adama

J. I.

, & Ogunjobi

J. O.

(2018). Financial sector and manufacturing sector performance: Evidence from Nigeria. Investment Management and Financial Innovations , 15(3), 35–48.

Asaleye

A. J.

, Maimako

R. F.

, Inegbedion

, Lawal

A. I.

, & Ogundipe

A. A.

(2021). Real exchange rate and manufacturing performance in Nigeria. Academic Journal of Interdisciplinary Studies , 10(2), 279–279.

Asteriou

, & Hall

S. G.

(2007). Applied econometrics: A modern approach . Palgrave McMillan.

Brox

J. A.

, & Fader

C. A.

(2005). Infrastructure investment and Canadian manufacturing productivity. Applied Economics , 37(11), 1247–1256.

Chakravarty

, & Mitra

(2009). Is industry still the engine of growth? An econometric study of the organized sector employment in India. Journal of Policy Modeling , 31(1), 22–35.

Dantama

Y. U.

, Abdullahi

Y. Z.

, & Inuwa

(2012). Energy consumption-economic growth nexus in Nigeria: An empirical assessment based on ARDL bound test approach. European Scientific Journal , 8(12), 141–157.

10.

Dasgupta

, & Singh

(2007). Manufacturing, services and premature deindustrialization in developing countries: A Kaldorian analysis. In Advancing development (pp. 435–454). Palgrave Macmillan.

11.

Dickey

D. A.

, & Fuller

W. A.

(1981). Likelihood ratio statistics for autoregressive time series with a unit root. Econometrica: Journal of the Econometric Society , 49(4), 1057–1072.

12.

Ekpo

U. N.

(2014). Nigeria industrial policies and industrial sector performance: Analytical exploration. IOSR Journal of Economics and Finance (IOSR-JEF) , 3(4), 1–11.

13.

Elhiraika

A. B.

, Aboubakar

, & Muhammad

(2014). Promoting manufacturing to accelerate economic growth and reduce growth volatility in Africa. The Journal of Developing Areas , 48(2), 1–20.

14.

Emmanuel

F. O.

, & Oladiran

O. I.

(2015). Effect of government capital expenditure on manufacturing sector output in Nigeria. Business and Economic Research , 5(2), 136–152.

15.

Fagerberg

, & Verspagen

(1999). Modern capitalism in the 1970s and 1980s. In Setterfield

(Ed.), Growth, employment and inflation (pp. 113–126). Palgrave Macmilan. https://link-springer-com-443.web.bisu.edu.cn/chapter/10.1007/978-1-349-27393-5_9

16.

Felipe

(1998). The role of the manufacturing sector in Southeast Asian development: A test of Kaldor’s first law. Journal of Post Keynesian Economics , 20(3), 463–485.

17.

Foellmi

, & Zweimüller

(2011). Exclusive goods and formal-sector employment. American Economic Journal: Macroeconomics , 3(1), 242–272.

18.

Forest

J. J.

, & Turner

(2013). Alternative estimators of cointegrating parameters in models with nonstationary data: An application to US export demand. Applied Economics , 45(5), 629–636.

19.

Ghani

, & O’Connell

S. D.

(2014). Can service be a growth escalator in low-income countries? (Policy Research Working Paper 6971). World Bank. http://hdl.handle.net/10986/19352

20.

Güçlü

(2013). Manufacturing and regional economic growth in Turkey: A spatial econometric view of Kaldor’s laws. European Planning Studies , 21(6), 854–866.

21.

Hammed

Y. S.

, & Arawomo

(2020). Oil price shock, fiscal policy and manufacturing sector in Nigeria: Evidence from SVAR. African Journal of Economic Review , 8(3), 66–83.

22.

Haraguchi

, Cheng

C. F. C.

, & Smeets

(2017). The importance of manufacturing in economic development: Has this changed? World Development , 93, 293–315. https://doi.org/10.1016/j.worlddev.2016.12.013

23.

Herman

(2016). The importance of the manufacturing sector in the Romanian economy. Procedia Technology , 22, 976–983. https://doi.org/10.1016/j.protcy.2016.01.121

24.

Ibrahim

M. H.

(2005). Sectoral effects of monetary policy: Evidence from Malaysia. Asian Journal Economic Journal , 19(1), 83–102.

25.

Idoko

C. U.

, & Taiga

U. U.

(2018). Effect of foreign direct investment (FDI) on manufacturing output in Nigeria (1981–2016). Advances in Social Sciences Research Journal , 5(5), 181–197.

26.

Johansen

, & Juselius

(1990). Maximum likelihood estimation and inference on cointegration—With applications to the demand for money. Oxford Bulletin of Economics and Statistics , 52(2), 169–210.

27.

Kaldor

(1966). Causes of the slow rate of economic growth of the United Kingdom: An inaugural lecture . Cambridge University Press.

28.

Kathuria

, & Natarajan

R. R.

(2013). Is manufacturing an engine of growth in India in the post-nineties? Journal of South Asian Development , 8(3), 385–408.

29.

Keho

(2018). Manufacturing and economic growth in ECOWAS countries: A test of Kaldor’s First Law. Modern Economy , 9(5), 897–906.

30.

Kripfganz

, & Schneider

D. C.

(2018). ardl: Estimating autoregressive distributed lag and equilibrium correction models [Paper presentation]. London Stata Conference 2018, No. 09. Stata Users Group. http://repec.org/usug2018/uk18_Kripfganz.pdf

31.

Liew

V. K.

, Rowland

, Hong

, Hui

J. K. S.

, Ab Rahim

R. B.

, & Ling

S. W. S.

(2018). Macroeconomic instability index and Malaysia economic performance. International Business Research , 11(3), 179–185.

32.

Marconi

, de Borja Reis

C. F.

, & de Araújo

E. C.

(2016). Manufacturing and economic development: The actuality of Kaldor’s first and second laws. Structural Change and Economic Dynamics , 37, 75–89. https://doi.org/10.1016/j.strueco.2015.12.002

33.

MacKinnon

K. G.

(1991). Critical values for cointegration test. In Engle

R. F.

, & Granger

C. W. J.

(Eds). Long Run Economic Relationships: Readings in Cointegration . Oxford University Press.

34.

McCausland

W. D.

, & Theodossiou

(2012). Is manufacturing still the engine of growth? Journal of Post Keynesian Economics , 35(1), 79–92.

35.

Mesagan

, & Ezeji

A. C.

(2016). The role of social and economic infrastructure in manufacturing sector performance in Nigeria (MPRA Paper 78310). University Library of Munich, Germany. https://ideas.repec.org/p/pra/mprapa/78310.html

36.

Moustafa

(2006). Does income inequality effect the composition of growth: The case of Egypt (Working Paper). University of Massachusetts, Amherst, MA, 1–33.

37.

Ndikumana

(2008). Can macroeconomic policy stimulate private investment in South Africa? New insights from aggregate and manufacturing sector-level evidence. Journal of International Development: The Journal of the Development Studies Association , 20(7), 869–887.

38.

Necmi

(1999). Kaldors growth analysis revisited. Applied Economics , 31(5), 653–660.

39.

Oladipo

O. A.

, Iyoha

, Fakile

S. A.

, Asaleye

A. J.

, & Eluyela

F. D.

(2019). Do government taxes have implications on manufacturing sector output? Evidence from Nigeria. Journal of Management Information and Decision Sciences , 22(3), 181–190.

40.

Ono

(2001). Determinants of trade credit in the Japanese sector. Journal of the Japanese and International Economies , 15(2), 160–177.

41.

Othman

A .N.

, & Masih

(2015). Do profit and loss sharing (PLS) deposits also affect PLS financing? Evidence from Malaysia based on DOLS, FMOLS and system GMM techniques (Working paper no. 65224 ) Islamic finance at INCEIF, Lorong Universiti A, 59100 Kuala Lumpur, Malaysia. https://mpra.ub.uni-muenchen.de/65224/

42.

Palma

(2005). Four sources of deindustrialization and a new concept of the Dutch Disease. Beyond Reforms: Structural Dynamics and Macroeconomic Vulnerability , 3(5), 71–116.

43.

Panopoulou

, & Pittis

(2004). A comparison of autoregressive distributed lag and dynamic OLS cointegration estimators in the case of a serially correlated cointegration error. The Econometrics Journal , 7(2), 585–617.

44.

Pata

U. K.

, & Zengin

(2020). Testing Kaldor’s growth laws for Turkey: New evidence from symmetric and asymmetric causality methods. Çankırı Karatekin Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi , 10(2), 713–729.

45.

Pesaran

M. H.

, Shin

, & Smith

R. J.

(2001). Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics , 16(3), 289–326.

46.

Phillips

P. C. B.

, & Perron

(1988). Testing for a unit root in time series regression. Biometrika , 75(2), 335–346.

47.

Pons-Novell

, & Viladecans-Marsal

(1999). Kaldor’s laws and spatial dependence: Evidence for the European regions. Regional Studies , 33(5), 443–451.

48.

Quoreshi

A. M. M.

, & Stone

T. A.

(2019). Do global value chains make firms more vulnerable to trade shocks?—Evidence from manufacturing firms in Sweden. Journal of Risk and Financial Management , 12(3), 151.

49.

Rodrik

(2016). Premature deindustrialization. Journal of Economic Growth , 21(1), 1–33.

50.

Rodrik

(2017). Premature deindustrialization in the developing world. Frontiers of Economics in China , 12(1), 1–7.

51.

Rowthorn

, & Ramaswamy

(1997). Deindustrialization: Causes and implications (pp. 1–38). International Monetary Fund.

52.

Rydzck

(2013). Trade effects of income inequality within and between countries (Working Paper, 1–30). https://www.econ.uzh.ch/dam/jcr:00000000-7972-abe4-0000-000010b092a0/wprydzek.pdf

53.

, & Yao

(2017). Manufacturing as the key engine of economic growth for middle-income economies. Journal of the Asia Pacific Economy , 22(1), 47–70.

54.

Szirmai

, & Verspagen

(2015). Manufacturing and economic growth in developing countries, 1950–2005. Structural Change and Economic Dynamics , 34, 46–59. https://doi.org/10.1016/j.strueco.2015.06.002

55.

Thirlwall

A. P.

(1983). A plain man’s guide to Kaldor’s growth laws. Journal of Post-Keynesian Economics , 5(3), 345–358.

56.

Tkalec

, & Vizek

(2009). The impact of macroeconomic policies on manufacturing production in Croatia. Privredna Kretanja i Ekonomska Politika , 19(121), 61–93.

57.

Tybout

J. R.

(2000). Manufacturing firms in developing countries: How well do they do, and why? Journal of Economic Literature , 38(1), 11–44.

58.

Varela

, Ghosh

, & Rahardha

(2012). How the macroeconomic environment and investment climate have affected the manufacturing sector? World Bank. https://ideas.repec.org/p/wbk/wboper/26717.html

59.

Wells

, & Thirlwall

A. P.

(2003). Testing Kaldor’s growth laws across the countries of Africa. African Development Review , 15(2–3), 89–105.

Sources of Instability in Nigeria’s Manufacturing Performance: 1981–2018

Abstract

Keywords

I. Introduction

II. Literature Review

III. Performance of the Nigerian Manufacturing Sector

Growth in Manufacturing Value Added by Selected Country (Percentage Value).

Unit Root Tests

Model Specification and Estimation Strategy

Data and Sources

Description of Variables.

Preliminary Analysis

Table 3.

Summary Descriptive Statistics.

Summary of Unit Root Tests.

Main Results

Bounds Test of Cointegration.

Estimated Long-Run Model.

Estimated Short-Run Model.

Summary of Post-Estimation Diagnostics.

Footnotes

Declaration of Conflicting Interests

Funding

ORCID iD

References