**4. Sudan: economic, social, and environmental dimensions of development**

We used the framework set in **Figure 1** to explain how social and environmental indicators have affected development in its economic dimension as commonly measured by GDP. Variable selection is necessitated by availability of data, which is collected from the World Bank, World Development Indicators (WDI), and World *Economic, Social, and Environmental Dimensions of Development in Sudan DOI: http://dx.doi.org/10.5772/intechopen.90752*

Bank 2018 [6] and complemented with other sources. The study variables are defined as follows:

Gross domestic product (GDP): GDP is the value of all goods and services produced in the economy expressed in current US dollars and stands for economic growth.

Domestic investment (INV): INV is measured by gross capital formation consisting of additions to the fixed assets of the economy plus net changes in the level of inventories. Fixed assets include land improvements (fences, ditches, drains, and so on); plant, machinery, and equipment purchases; and the construction of roads, railways, and the like, including schools, offices, hospitals, private residential dwellings, and commercial and industrial buildings. Inventories are stocks of goods held by firms to meet temporary or unexpected fluctuations in production or sales and work in progress.

Unemployment is represented by the total youth unemployment as percentage of total labor force ages (15–24), and it refers to the share of the labor force ages 15–24 without work but available for and seeking employment.

Life expectancy at birth indicates the number of years a newborn would live if prevailing patterns of mortality at the time of its birth were to stay the same throughout its life.

Average school enrollment (EDU): EDU is measured by net enrollment rate which is the ratio of children of official school age who are enrolled in school to the population of the corresponding official school age. The World Bank acknowledges that primary education provides children with basic reading, writing, and mathematics skills. These are basics for any progression to secondary and tertiary education.

Access to drinking water (ASW): Access to safe drinking water is measured by people using at least basic drinking water services (% of population). It encompasses both people using basic water services and those using safely managed water services. Basic drinking water services are defined as drinking water from an improved source, provided collection time is not more than 30 minutes for a round trip. Improved water sources include piped water, boreholes or tube wells, protected dug wells, protected springs, and packaged or delivered water.

Access to sanitation facilities (ASF): It is represented by people using safely managed sanitation services (% of population) defined as the percentage of people using improved sanitation facilities that are not shared with other households and where excreta are safely disposed of in situ or transported and treated off-site. Improved sanitation facilities include flush/pour flush to piped sewer systems, septic tanks, or pit latrines: ventilated improved pit latrines, compositing toilets, or pit latrines with slabs.

Access to electricity is measured by the average percentage of population with access to electricity.

Carbon dioxide emissions as defined by the World Bank are those stemming from the burning of fossil fuels and the manufacture of cement. They include carbon dioxide produced during consumption of solid, liquid, and gas fuels and gas flaring. CO2 emissions are measured in metric tons per capita (CO2P).

Trade openness is measured as the sum of exports and imports of goods and services as percentage of gross domestic product.

#### **5. Empirical analysis**

#### **5.1 Descriptive statistical analysis**

**Table 1** presents the descriptive statistics of the study variables. From the Jarque-Bera (J-B) and associated prob. values, all variables look normally

resources, both renewable and nonrenewable. Even this pattern has been a major dragger to economic growth in the sense of the so-called resource curse hypothesis. These facts and accounts are our rationale to model GDP growth against social and

*Perspectives on Economic Development - Public Policy, Culture, and Economic Development*

**3. Analytical framework: from social and environmental performance**

**Figure 1** shows a proposed analytical framework in which social and environmental indicators, together with investment and trade openness, are assumed to explain economic growth represented by the current gross domestic product (GDP). For a low-income country such as Sudan, social and environmental dimensions of development are more expected to lead economic dimension of development rather than economic leading to socio-environmental improvements. The framework also assumes that selected all economic, social, and environmental measures to affect economic development positively. In terms of causation, the analytical framework presumes that the causation runs from social and environmental performance to economic development as narrowly measured by GDP growth.

environmental performance indicators, rather than the other way around.

**4. Sudan: economic, social, and environmental dimensions of**

*Economic, social, and environmental dimensions of development.*

We used the framework set in **Figure 1** to explain how social and environmental

indicators have affected development in its economic dimension as commonly measured by GDP. Variable selection is necessitated by availability of data, which is collected from the World Bank, World Development Indicators (WDI), and World

**development**

**Figure 1.**

**66**

**to economic performance**

distributed expect education, access to drinking water, and access to electricity. The highest kurtosis is associated with GDP followed by access to drinking water and access to electricity. As for average rainfall, it has been reported that summer monthly precipitation over the Sahel is not normally distributed.

From the correlation matrix in **Table 2** noticeably, the GDP highly positively correlates with life expectancy at birth, education, access to drinking water, and access to electricity. Life expectancy and education both positively correlate with access to drinking water and access to electricity.

We conducted a graphical analysis of the study variables. **Figure 2** shows upward trend for current GDP, while **Figure 3** depicts erratic trend of investment. Also youth unemployment experiences high fluctuations as depicted in **Figure 4**. **Figure 5** indicates upward sloping life expectancy at birth over time. School enrolment shows upward trend with some small fluctuations as shown in **Figure 6**. Access to sanitation facilities runs through a period of upward trend and downward trend and more recently started to trend upward as depicted in **Figure 7**. Access to drinking water showed slow increase at the beginning of our time series but started to increase sharply after the year 2004 as shown in **Figure 8**. Access to electricity


#### **Table 1.**

*Basic statistics.*


shows more erratic trend over time and only after 2002 started to show steady increase but with a drop in 2009 and a hike in 2014 as in **Figure 9**. Carbon dioxide emissions experienced a declining trend from 1974 until 1993 and then started to

*Economic, Social, and Environmental Dimensions of Development in Sudan*

*DOI: http://dx.doi.org/10.5772/intechopen.90752*

**Figure 2.** *Log of GDP.*

**Figure 3.** *Log of investment.*

**Figure 4.**

**69**

*Log of youth unemployment.*

**Table 2.** *Correlation matrix.*

*Economic, Social, and Environmental Dimensions of Development in Sudan DOI: http://dx.doi.org/10.5772/intechopen.90752*

**Figure 2.** *Log of GDP.*

distributed expect education, access to drinking water, and access to electricity. The highest kurtosis is associated with GDP followed by access to drinking water and access to electricity. As for average rainfall, it has been reported that summer

*Perspectives on Economic Development - Public Policy, Culture, and Economic Development*

From the correlation matrix in **Table 2** noticeably, the GDP highly positively correlates with life expectancy at birth, education, access to drinking water, and access to electricity. Life expectancy and education both positively correlate with

We conducted a graphical analysis of the study variables. **Figure 2** shows upward trend for current GDP, while **Figure 3** depicts erratic trend of investment. Also youth unemployment experiences high fluctuations as depicted in **Figure 4**. **Figure 5** indicates upward sloping life expectancy at birth over time. School enrolment shows upward trend with some small fluctuations as shown in **Figure 6**. Access to sanitation facilities runs through a period of upward trend and downward trend and more recently started to trend upward as depicted in **Figure 7**. Access to drinking water showed slow increase at the beginning of our time series but started to increase sharply after the year 2004 as shown in **Figure 8**. Access to electricity

**GDP INV YUN LE EDU ASF ASW ELC CO2P TOP**

**GDP INV YUN LE EDU ASF ASW ELC CO2P TOP**

Mean 2.60E+10 19.09 27.85 57.20 28.64 25.35 45.07 30.71 0.24 26.80 Median 1.24E+10 16.34 27.82 56.19 25.07 25.35 42.73 29.58 0.22 25.46 Maximum 1.17E+11 37.19 28.95 64.99 48.41 27.70 59.27 44.90 0.37 47.58 Minimum 2.44E+09 7.29 26.39 51.74 18.67 22.50 40.62 23.00 0.10 11.09 Std. dev. 2.91E+10 7.27 0.59 3.71 8.92 1.45 5.59 4.37 0.09 9.53 Skewness 1.60 0.67 0.03 0.51 0.93 0.15 1.52 0.87 0.28 0.19 Kurtosis 4.43 2.49 2.51 2.12 2.48 1.82 3.93 3.93 1.54 2.31 J-B 24.52 4.14 0.50 3.67 7.44 2.97 20.32 7.81 4.91 1.25 Prob. 0.000 0.126 0.780 0.160 0.024 0.226 0.000 0.020 0.086 0.535 Obs. 48 48 48 48 48 48 48 48 48 48

monthly precipitation over the Sahel is not normally distributed.

access to drinking water and access to electricity.

**Table 1.** *Basic statistics.*

**Table 2.** *Correlation matrix.*

**68**

GDP 1.00

INV 0.09 1.00

YUN 0.45 0.19 1.00

LE 0.90 0.22 0.29 1.00

EDU 0.93 0.10 0.35 0.97 1.00

ASF 0.44 0.06 0.31 0.48 0.54 1.00

ASW 0.98 0.06 0.39 0.91 0.95 0.45 1.00

ELC 0.83 0.23 0.24 0.93 0.89 0.32 0.87 1.00

CO2P 0.63 0.26 0.54 0.44 0.60 0.67 0.62 0.36 1.00

TOP 0.04 0.22 0.22 0.10 0.15 0.65 0.03 0.09 0.44 1.00

**Figure 3.** *Log of investment.*

*Log of youth unemployment.*

shows more erratic trend over time and only after 2002 started to show steady increase but with a drop in 2009 and a hike in 2014 as in **Figure 9**. Carbon dioxide emissions experienced a declining trend from 1974 until 1993 and then started to

rest of our time frameworks in **Figure 11**. These time trends of the economic, social, and environmental dimensions of development in Sudan reflect various episodes of war and environmental and climate change along generally failed macroeconomic policies to improve development with its multifacet and interacting dimensions.

*Economic, Social, and Environmental Dimensions of Development in Sudan*

*DOI: http://dx.doi.org/10.5772/intechopen.90752*

**Figure 8.**

**Figure 9.**

**Figure 10.**

**71**

*Log of carbon dioxide emissions per capita.*

*Log of access to electricity.*

*Log of access to drinking water.*

**Figure 5.** *Log of life expectancy at birth.*

**Figure 6.** *Log of primary school enrolment.*

**Figure 7.** *Log of access to sanitation facilities.*

increase over the rest of our time series frame as in **Figure 10**. Trade openness runs through relatively stable trend over the period 1970–1983, a massive decline from 1984 to 1993, an increase over the period 1994–2006, and then a decline over the

rest of our time frameworks in **Figure 11**. These time trends of the economic, social, and environmental dimensions of development in Sudan reflect various episodes of war and environmental and climate change along generally failed macroeconomic policies to improve development with its multifacet and interacting dimensions.

**Figure 8.** *Log of access to drinking water.*

**Figure 9.** *Log of access to electricity.*

**Figure 10.** *Log of carbon dioxide emissions per capita.*

increase over the rest of our time series frame as in **Figure 10**. Trade openness runs through relatively stable trend over the period 1970–1983, a massive decline from 1984 to 1993, an increase over the period 1994–2006, and then a decline over the

*Perspectives on Economic Development - Public Policy, Culture, and Economic Development*

**Figure 5.**

**Figure 6.**

**Figure 7.**

**70**

*Log of primary school enrolment.*

*Log of access to sanitation facilities.*

*Log of life expectancy at birth.*

However, a nonstationary series possessing a stochastic unit root can be differenced once to become stationary. Establishing stationarity of macroeconomic series is necessary for reliable econometric estimations and causality analysis tests and since most macroeconomic series are in fact not stationary. Stationary of time series included in this study is tested through the conventional augmented Dickey-Fuller (ADF) and Phillips-Perron (PP) unit root tests, complimented by Kwiatkowski-Phillips-Schmidt-Shin (KPSS [8]). The ADF test takes into account only the presence of autocorrelation in the series, but the PP test considers also the hypothesis of the presence of a heteroskedasticity dimension in the time series and Kwiatkowski-Phillips-Schmidt-Shin (KPSS [8]). In literature, tests designed following the null hypothesis that a series is I(1) have low power to reject the null. Therefore, KPSS is sometimes used along the widely used ADF and PP tests to have robust results. The findings from the ADF, PP, and KPSS tests are reported in **Table 3**. Investment and unemployment are found to be stationary at level I(0) and first difference I(1)

*Economic, Social, and Environmental Dimensions of Development in Sudan*

*DOI: http://dx.doi.org/10.5772/intechopen.90752*

Combined results from these three tests indicate all series are integrated to the order i.e., I(1) but some of them are stationary at level I(0), i.e., all series are I(0)/I (1). As a result, we choose to use the autoregressive distributed lag (ARDL) bound test for cointegration because one of its main advantages is that it does not impose a

**Variable ADF PP KPSS Order of integration**

**Stat. 5% cri. value Stat. 5% cri. value Stat.** L(GDP) 0.380 2.925 0.400 2.925 0.799 Nonstationary *<sup>Δ</sup>L(GDP) 6.349\* 2.927 6.349\* 2.927 0.136\* Stationary I(1)* L(INV) 3.048\* 2.925 2.897 2.925 0.450\* Stationary I(0) *<sup>Δ</sup>L(INV) 6.498\* 2.928 12.465\* 2.927 0.500 Stationary I(1)* L(YUN) 2.555 2.927 5.338\* 2.925 0.201\* Stationary I(0) *<sup>Δ</sup>L(YUN) 8.931\* 2.928 16.341\* 2.927 0.020\* Stationary I(1)* L(LE) 2.102 2.931 2.679 2.925 0.891 Nonstationary *<sup>Δ</sup>L(LE) 2.606 2.929 5.747\* 2.927 0.632 Stationary I(1)* L(EDU) 0.025 2.928 0.282 2.925 0.788 Nonstationary *<sup>Δ</sup>L(EDU) 10.722\* 2.928 11.275\* 2.927 0.239\* Stationary I(1)* L(ASF) 1.941 2.931 2.131 2.925 0.325\* Stationary I(0) *<sup>Δ</sup>L(ASF) 1.916 2.931 7.154\* 2.927 0.131\* Stationary I(1)* L(ASW) 2.643 2.925 2.487 2.925 0.707 Stationary I(0) *<sup>Δ</sup>L(ASW) 3.243\* 2.928 5.924\* 2.927 0.569 Stationary I(1)* L(ELC) 0.050 2.927 0.259 2.925 0.858 Nonstationary *<sup>Δ</sup>L(ELC) 6.829\* 2.927 13.650\* 2.927 0.319\* Stationary I(1)* L(CO2P) 1.097 2.927 1.391 2.925 0.279\* Stationary I(0) *<sup>Δ</sup>(CO2P) 9.243\* 2.927 9.167\* 2.927 0.254\* Stationary I(1)* L(TOP) 1.920 2.925 1.888 2.925 0.133\* Stationary I(0) *<sup>Δ</sup> L(TOP) 8.488\* 2.927 8.370\* 2.927 0.084\* Stationary I(1) Note: The ADF and PP unit root tests employ null hypothesis with the series that has a unit root against the alternative of stationary. The null hypothesis for the KPSS assumes that the variable is stationary. KPSS critical value is 0.463.*

while the first differencing makes all variables stationary.

*\*indicates significance at 5% level.*

**Table 3.**

**73**

*Unit root test results.*

#### **5.2 Empirical investigation**

#### *5.2.1 Econometric methods*

This study is empirical and quantitative, using statistical and econometric methods. Sound empirical studies on economic, social, and environmental dimensions of development need to be based on clear theoretical framework, rigorous methodology, and reliable data. Empirical quantitative studies using dynamic econometric methods on these relationships are rare, although empiricalquantitative research programs in all socioeconomic and environmental issues are usually more rigorous [7]. The study is very selective on variables included, which is necessitated by data availability and possible theoretical links. The study presumes that economic growth is affected by social and environmental progress indicators rather than the other way around. The study covers the period 1970–2017 with annual time series data on all of its variables. Economic growth represented by current GDP is treated as the dependent variable, and social and environmental indicators are the independent variables. Trade openness is included as a control variable and represents the exposure of Sudan economy to international shocks. A general log linear model to capture the complexity of economic, social, and environmental dimensions of development in Sudan is written as:

$$\begin{aligned} L(\text{GDP}) &= \alpha + \beta\_1 L(\text{INV}) + \beta\_2 L(\text{UNE}) + \beta\_3 L(\text{LE}) + \beta\_4 L(\text{EDU}) + \beta\_5 L(\text{ASF}) \\ &+ \beta\_6 L(\text{ASW}) + \beta\_7 L(\text{ELC}) + \beta\_8 L(\text{CO2P}) + \beta\_9 L(\text{TOP}) + \beta\_{10} D\text{UIM} + \mu \end{aligned} \tag{1}$$

where DUM stands for dummy that is 0 in 1978, 1997, and 2011 and 1 otherwise. These years are judged to represent breaks in Sudan economy as years of the first ever devaluation of the national currency, imposition of sanction on Sudan by the United States, and secession of South Sudan, respectively.

#### *5.2.2 Stationary and cointegration of variables*

The first step is to use a preliminary statistic test to verify the stationarity for all variables. A time series is described as nonstationary if it has at least one of its moments (mean, variance, or covariance) as time independent.

#### *Economic, Social, and Environmental Dimensions of Development in Sudan DOI: http://dx.doi.org/10.5772/intechopen.90752*

However, a nonstationary series possessing a stochastic unit root can be differenced once to become stationary. Establishing stationarity of macroeconomic series is necessary for reliable econometric estimations and causality analysis tests and since most macroeconomic series are in fact not stationary. Stationary of time series included in this study is tested through the conventional augmented Dickey-Fuller (ADF) and Phillips-Perron (PP) unit root tests, complimented by Kwiatkowski-Phillips-Schmidt-Shin (KPSS [8]). The ADF test takes into account only the presence of autocorrelation in the series, but the PP test considers also the hypothesis of the presence of a heteroskedasticity dimension in the time series and Kwiatkowski-Phillips-Schmidt-Shin (KPSS [8]). In literature, tests designed following the null hypothesis that a series is I(1) have low power to reject the null. Therefore, KPSS is sometimes used along the widely used ADF and PP tests to have robust results. The findings from the ADF, PP, and KPSS tests are reported in **Table 3**. Investment and unemployment are found to be stationary at level I(0) and first difference I(1) while the first differencing makes all variables stationary.

Combined results from these three tests indicate all series are integrated to the order i.e., I(1) but some of them are stationary at level I(0), i.e., all series are I(0)/I (1). As a result, we choose to use the autoregressive distributed lag (ARDL) bound test for cointegration because one of its main advantages is that it does not impose a


*Note: The ADF and PP unit root tests employ null hypothesis with the series that has a unit root against the alternative of stationary. The null hypothesis for the KPSS assumes that the variable is stationary. KPSS critical value is 0.463. \*indicates significance at 5% level.*

#### **Table 3.** *Unit root test results.*

**5.2 Empirical investigation**

This study is empirical and quantitative, using statistical and econometric methods. Sound empirical studies on economic, social, and environmental dimensions of development need to be based on clear theoretical framework, rigorous methodology, and reliable data. Empirical quantitative studies using dynamic econometric methods on these relationships are rare, although empirical-

*Perspectives on Economic Development - Public Policy, Culture, and Economic Development*

quantitative research programs in all socioeconomic and environmental issues are usually more rigorous [7]. The study is very selective on variables included, which is necessitated by data availability and possible theoretical links. The study presumes that economic growth is affected by social and environmental progress indicators rather than the other way around. The study covers the period 1970–2017 with annual time series data on all of its variables. Economic growth represented by current GDP is treated as the dependent variable, and social and environmental indicators are the independent variables. Trade openness is included as a control variable and represents the exposure of Sudan economy to international shocks. A general log linear model to capture the complexity of economic, social, and envi-

*L GDP* ð Þ¼ *α* þ *β*1*L INV* ð Þþ *β*2*L UNE* ð Þþ *β*3*L LE* ð Þþ *β*4*L EDU* ð Þþ *β*5*L ASF* ð Þ

þ *β*6*L ASW* ð Þþ *β*7*L ELC* ð Þþ *β*8*L CO* ð Þþ 2*P β*9*L TOP* ð Þþ *β*10*DUM* þ *μ*

where DUM stands for dummy that is 0 in 1978, 1997, and 2011 and 1 otherwise. These years are judged to represent breaks in Sudan economy as years of the first ever devaluation of the national currency, imposition of sanction on Sudan by the

The first step is to use a preliminary statistic test to verify the stationarity for all variables. A time series is described as nonstationary if it has at least one of

(1)

ronmental dimensions of development in Sudan is written as:

United States, and secession of South Sudan, respectively.

its moments (mean, variance, or covariance) as time independent.

*5.2.2 Stationary and cointegration of variables*

**72**

*5.2.1 Econometric methods*

**Figure 11.** *Log of trade openness.* restrictive assumption that all variables should have the same integration order. Another advantage is that a dynamic error correction (EC) term can be derived from the ARDL through simple linear transformation. The error correction term shows the short-run dynamics with the long-run stable equilibrium without losing the long-run information.

#### *5.2.3 ARDL model specification and estimation*

An autoregressive distributed lag model bound test advanced by Pesaran and Smith [9], Pesaran and Shin [10] and with the bound test of Pesaran et al. [11] is used to investigate cointegration and the short-run dynamics and long-run equilibrium of GDP as the dependent variable and social and environmental indicators as explanatory variables. An unrestricted ARDL model on the basis of Eq. (1) is specified as follows:

$$\begin{split} \Delta L(GDP)\_{t} &= a + \sum\_{i=1}^{p} \beta\_{1i} L(GDP)\_{t-1} \\ &+ \sum\_{i=0}^{p} \Delta \beta\_{2i} L(INV)\_{t-i} + \sum\_{i=0}^{p} \beta\_{3i} \Delta L(UNE)\_{t-i} \\ &+ \sum\_{i=0}^{p} \beta\_{4i} \Delta L(LE)\_{t-i} + \sum\_{i=0}^{p} \beta\_{5} \Delta L(EOV)\_{t-i} \\ &+ \sum\_{i=0}^{p} \beta\_{6} \Delta L(ASF)\_{t-i} + \sum\_{i=0}^{p} \beta\_{7} \Delta L(ASW)\_{t-i} \\ &+ \sum\_{i=0}^{p} \beta\_{8} \Delta L(ELC)\_{t-i} + \sum\_{i=0}^{p} \beta\_{9} \Delta L(CO2P)\_{t-i} \\ &+ \sum\_{i=0}^{p} \beta\_{10} \Delta L(TOP)\_{t-i} + \beta\_{10} L(GDP)\_{t-1} + \beta\_{11} L(INV)\_{t-1} \\ &+ \beta\_{12} L(UNE)\_{t-1} + \beta\_{13} L(LE\_{t-1} + \beta\_{14} L(EDU)\_{t-1} \\ &+ \beta\_{15} L(ASF)\_{t-1} + \beta\_{16} L(TOP)\_{t-1} + \beta\_{20} UIDM\_t + EC\_t + \mu\_t \\ &+ \beta\_{18} L(CO2P)\_{t-1} + \beta\_{19} L(TOP)\_{t-1} + \beta\_{20} DUDM\_t + EC\_t + \mu\_t \end{split} (2)$$

variable is treated as dependent while all other variables are independent. The results show that there at least eight cointegrated forms as summarized in **Table 7**. Thus, the results of the bound test cointegration confirm the existence of a long-run equilibrium relationship between economic growth measured by GDP in relation to the set of social and environmental indicators and other covariates included as

*LR, sequential modified LR test statistic (each test at 5% level); FPE, final prediction error; AIC, Akaike information criterion; SC,*

**F-bound test Null hypothesis: no level relationship**

*Economic, Social, and Environmental Dimensions of Development in Sudan*

**Test statistic Value Significance I(0) I(1)** F-statistic 7.47 10% 1.8 2.8 K 9 5% 2.04 2.08

**Lag LL LR FPE AIC SC HQ** 467.371 NA 7.03e-22 20.328 19.926 20.178 791.748 490.169 3.63e-26 30.300 25.884\* 28.654 945.800 *164.322\*#* 6.50e-27 32.702 24.271 29.560 1132.692 116.289 1.44e-27\* 36.564\* 24.118 31.924\*

2.5% 2.24 3.35 1% 2.5 3.68

The main ARDL results are summarized in **Table 6**.

long run, although their coefficients are positive in two time frames.

Youth unemployment is the most hurting to economic growth in Sudan. The

We then turn to investigate how economic growth and social and environmental indicators interact in the short run and the long run. For this purpose, an ARDL to be estimated is chosen out of 39,366 models (2, 2, 0, 2, 0, 2, 2, 1, 0, 2) on the basis of

The ARDL short-run dynamics and long-run equilibrium results are summarized

ARDL short-run dynamics and error correction (EC) results are summarized in

The ARDL model shows that in the short run, investment has positive effect on GDP but with 1-year time lag. Life expectancy at birth has a negative effect on GDP with the highest elasticity coefficient of 16.45. Access to sanitation facilities exerts negative effect on GDP, while access to drinking water exerts positive effect on GDP although at 1-year time lag. Trade openness also has a negative effect on GDP. In the long run, GDP growth is positively and highly significantly affected by life expectancy at birth, access to sanitation facilities, and trade openness, respectively, while GDP is found to be negatively affected by youth unemployment followed by access to drinking water and investment. These long-run effects indicate that investments in physical capital and drinking water services have either been insufficient or ineffective in promoting economic growth in Sudan. Carbon dioxide emissions have no significant effect on economic growth in both the short and

reported in **Table 4**.

in **Table 7**.

**Table 8**.

**75**

**Table 4.**

*\**

*# at 5% level*

**Table 5.**

*Bound test cointegration: GDP dependent variable.*

*DOI: http://dx.doi.org/10.5772/intechopen.90752*

*Schwarz information criterion; HQ, Hannan-Quinn information criterion.*

criteria reported in **Table 5**.

*Indicates lag order selected by the criterion;*

*VAR lag order selection criteria.*

All variables and abbreviations are as defined above. The parameter *p* is the lag length, *Δ* is the difference operator, and EC is the ARDL error correction term. Eq. (2) can be estimated through the OLS to explore the long-run relationship of the model variables by performing an F-test statistics for the joint significance of the lagged-level variables. The null hypothesis of no cointegration (i.e., no long-run equilibrium relationship between the study variables) in Eq. (2) is: *H*<sup>0</sup> : *β*<sup>1</sup> ¼ *β*<sup>2</sup> ¼ *β*<sup>3</sup> ¼ *β*<sup>4</sup> ¼ *β*<sup>5</sup> ¼ *β*<sup>6</sup> ¼ *β*<sup>7</sup> ¼ *β*<sup>8</sup> þ *β*<sup>9</sup> ¼ *β*<sup>10</sup> ¼ 0, against the alternative hypothesis of the existence of cointegration that: *H*<sup>1</sup> : *β*<sup>11</sup> 6¼ *β*<sup>12</sup> 6¼ *β*<sup>13</sup> 6¼ *β*<sup>14</sup> 6¼ *β*<sup>15</sup> 6¼ *β*<sup>16</sup> 6¼ *β*<sup>17</sup> 6¼ *β*<sup>18</sup> 6¼ *β*<sup>19</sup> 6¼ *β*<sup>20</sup> 6¼ 0.

The decision rule for the existence of cointegration in the bound testing approach according to Pesaran and Shin (1999) is two sets of critical values for the F-statistic: the lower bound where all variables are cointegrated of the order I(0) and the upper bound where all variables are cointegrated of the order I(1). If the Fstatistic lies below the lower bound value, the conclusion is no cointegration, and if the F-statistic is found to be above the upper bound value, then cointegration exists, whereas if the F-statistic falls between the upper bound and the lower bound, then the test is inconclusive. The ARDL bound test is performed on Eq. (2), where each


*Economic, Social, and Environmental Dimensions of Development in Sudan DOI: http://dx.doi.org/10.5772/intechopen.90752*

#### **Table 4.**

restrictive assumption that all variables should have the same integration order. Another advantage is that a dynamic error correction (EC) term can be derived from the ARDL through simple linear transformation. The error correction term shows the short-run dynamics with the long-run stable equilibrium without losing

*Perspectives on Economic Development - Public Policy, Culture, and Economic Development*

An autoregressive distributed lag model bound test advanced by Pesaran and Smith [9], Pesaran and Shin [10] and with the bound test of Pesaran et al. [11] is used to investigate cointegration and the short-run dynamics and long-run equilibrium of GDP as the dependent variable and social and environmental indicators as explanatory variables. An unrestricted ARDL model on the basis of Eq. (1) is

*p*

*i*¼0

*p*

*i*¼0

*p*

*i*¼0

*p*

*i*¼0

<sup>þ</sup> *<sup>β</sup>*12*L UNE* ð Þ*<sup>t</sup>*�<sup>1</sup> <sup>þ</sup> *<sup>β</sup>*13*L LE* ð Þ*<sup>t</sup>*�<sup>1</sup> <sup>þ</sup> *<sup>β</sup>*14*L EDU* ð Þ*<sup>t</sup>*�<sup>1</sup> <sup>þ</sup> *<sup>β</sup>*15*L ASF* ð Þ*<sup>t</sup>*�<sup>1</sup> <sup>þ</sup> *<sup>β</sup>*16*L ASW* ð Þ*<sup>t</sup>*�<sup>1</sup> <sup>þ</sup> *<sup>β</sup>*17*L ELC* ð Þ*<sup>t</sup>*�<sup>1</sup>

The decision rule for the existence of cointegration in the bound testing approach according to Pesaran and Shin (1999) is two sets of critical values for the F-statistic: the lower bound where all variables are cointegrated of the order I(0) and the upper bound where all variables are cointegrated of the order I(1). If the Fstatistic lies below the lower bound value, the conclusion is no cointegration, and if the F-statistic is found to be above the upper bound value, then cointegration exists, whereas if the F-statistic falls between the upper bound and the lower bound, then the test is inconclusive. The ARDL bound test is performed on Eq. (2), where each

*<sup>β</sup>*3*<sup>i</sup>*Δ*L UNE* ð Þ*<sup>t</sup>*�*<sup>i</sup>*

*<sup>β</sup>*5Δ*L EDU* ð Þ*<sup>t</sup>*�*<sup>i</sup>*

*<sup>β</sup>*7Δ*L ASW* ð Þ*<sup>t</sup>*�*<sup>i</sup>*

(2)

*<sup>β</sup>*9Δ*L CO* ð Þ <sup>2</sup>*<sup>P</sup> <sup>t</sup>*�*<sup>i</sup>*

*<sup>β</sup>*10Δ*L TOP* ð Þ*<sup>t</sup>*�*<sup>i</sup>* <sup>þ</sup> *<sup>β</sup>*10*L GDP* ð Þ*<sup>t</sup>*�<sup>1</sup> <sup>þ</sup> *<sup>β</sup>*11*L INV* ð Þ*<sup>t</sup>*�<sup>1</sup>

<sup>þ</sup> *<sup>β</sup>*18*L CO* ð Þ <sup>2</sup>*<sup>P</sup> <sup>t</sup>*�<sup>1</sup> <sup>þ</sup> *<sup>β</sup>*19*L TOP* ð Þ*<sup>t</sup>*�<sup>1</sup> <sup>þ</sup> *<sup>β</sup>*20*DUMt* <sup>þ</sup> *ECt* <sup>þ</sup> *<sup>μ</sup><sup>t</sup>*

All variables and abbreviations are as defined above. The parameter *p* is the lag length, *Δ* is the difference operator, and EC is the ARDL error correction term. Eq. (2) can be estimated through the OLS to explore the long-run relationship of the model variables by performing an F-test statistics for the joint significance of the lagged-level variables. The null hypothesis of no cointegration (i.e., no long-run equilibrium relationship between the study variables) in Eq. (2) is: *H*<sup>0</sup> : *β*<sup>1</sup> ¼ *β*<sup>2</sup> ¼ *β*<sup>3</sup> ¼ *β*<sup>4</sup> ¼ *β*<sup>5</sup> ¼ *β*<sup>6</sup> ¼ *β*<sup>7</sup> ¼ *β*<sup>8</sup> þ *β*<sup>9</sup> ¼ *β*<sup>10</sup> ¼ 0, against the alternative hypothesis of the existence of cointegration that: *H*<sup>1</sup> : *β*<sup>11</sup> 6¼ *β*<sup>12</sup> 6¼ *β*<sup>13</sup> 6¼ *β*<sup>14</sup> 6¼ *β*<sup>15</sup> 6¼ *β*<sup>16</sup> 6¼ *β*<sup>17</sup> 6¼

the long-run information.

specified as follows:

<sup>Δ</sup>*L GDP* ð Þ*<sup>t</sup>* <sup>¼</sup> *<sup>α</sup>* <sup>þ</sup><sup>X</sup>

þ X *p*

þ X *p*

þ X *p*

þ X *p*

þ X *p*

*β*<sup>18</sup> 6¼ *β*<sup>19</sup> 6¼ *β*<sup>20</sup> 6¼ 0.

**74**

*i*¼0

*i*¼0

*i*¼0

*i*¼0

*i*¼0

*5.2.3 ARDL model specification and estimation*

*p*

*i*¼1

*<sup>β</sup>*1*iL GDP* ð Þ*<sup>t</sup>*�<sup>1</sup>

<sup>Δ</sup>*β*2*iL INV* ð Þ*<sup>t</sup>*�*<sup>i</sup>* <sup>þ</sup><sup>X</sup>

*<sup>β</sup>*4*<sup>i</sup>*Δ*L LE* ð Þ*<sup>t</sup>*�*<sup>i</sup>* <sup>þ</sup><sup>X</sup>

*<sup>β</sup>*6Δ*L ASF* ð Þ*<sup>t</sup>*�*<sup>i</sup>* <sup>þ</sup><sup>X</sup>

*<sup>β</sup>*8Δ*L ELC* ð Þ*<sup>t</sup>*�*<sup>i</sup>* <sup>þ</sup><sup>X</sup>

*Bound test cointegration: GDP dependent variable.*


*\* Indicates lag order selected by the criterion;*

*# at 5% level*

*LR, sequential modified LR test statistic (each test at 5% level); FPE, final prediction error; AIC, Akaike information criterion; SC, Schwarz information criterion; HQ, Hannan-Quinn information criterion.*

#### **Table 5.**

*VAR lag order selection criteria.*

variable is treated as dependent while all other variables are independent. The results show that there at least eight cointegrated forms as summarized in **Table 7**. Thus, the results of the bound test cointegration confirm the existence of a long-run equilibrium relationship between economic growth measured by GDP in relation to the set of social and environmental indicators and other covariates included as reported in **Table 4**.

We then turn to investigate how economic growth and social and environmental indicators interact in the short run and the long run. For this purpose, an ARDL to be estimated is chosen out of 39,366 models (2, 2, 0, 2, 0, 2, 2, 1, 0, 2) on the basis of criteria reported in **Table 5**.

The main ARDL results are summarized in **Table 6**.

The ARDL short-run dynamics and long-run equilibrium results are summarized in **Table 7**.

ARDL short-run dynamics and error correction (EC) results are summarized in **Table 8**.

The ARDL model shows that in the short run, investment has positive effect on GDP but with 1-year time lag. Life expectancy at birth has a negative effect on GDP with the highest elasticity coefficient of 16.45. Access to sanitation facilities exerts negative effect on GDP, while access to drinking water exerts positive effect on GDP although at 1-year time lag. Trade openness also has a negative effect on GDP. In the long run, GDP growth is positively and highly significantly affected by life expectancy at birth, access to sanitation facilities, and trade openness, respectively, while GDP is found to be negatively affected by youth unemployment followed by access to drinking water and investment. These long-run effects indicate that investments in physical capital and drinking water services have either been insufficient or ineffective in promoting economic growth in Sudan. Carbon dioxide emissions have no significant effect on economic growth in both the short and long run, although their coefficients are positive in two time frames. Youth unemployment is the most hurting to economic growth in Sudan. The

coefficient of the error correction term estimated at 0.53 is highly significant confirming cointegration of the study variables and average speed of adjustment to equilibrium in the long run in response to the short-run shocks of the model variables. Results of the diagnostic tests show that the estimated ARDL suffers none of the conventional econometric problems associated with time series data. Thus, the estimated model is stable and robust and significantly captures the behavior of the association between economic growth and social and environmental progress indicators.


P(0.000); AIC = 1.422; SC = 0.468; HQ = 1.064; DW = 2.15.


As evident from **Figures 12** and **13**, all plots of cumulative sum (CUSUM) and cumulative sum of squares (CUSUMSQ) statistics of the recursive residuals are well within the critical bounds, implying that the coefficients in the error correction model of the

**Variable Coefficient Std. error t-statistic Prob.** L(INV) 1.03 0.228 4.537 0.000\*\*\* L(YUN) 5.79 2.465 2.347 0.028\*\* L(LE) 19.44 4.274 4.5480 0.000\*\*\* L(EDU) 0.53 0.842 0.629 0.536 L(ASF) 6.95 2.507 2.771 0.011\*\* L(ASW) 2.91 1.471 1.978 0.061\* L(ELC) 1.15 0.716 1.601 0.124 L(CO2P) 0.04 0.236 0.164 0.871 L(TOP) 0.80 0.257 3.114 0.005\*\*\* <sup>C</sup> 52.28 19.266 2.713 0.013\*\*\* EC = L(GDP) (1.03 L(INV) 5.79 L(YUN) + 19.44 L(LE) + 0.53 L(EDU) + 6.95 L(ASF) 2.91 L

*Economic, Social, and Environmental Dimensions of Development in Sudan*

L(TOP) 52.28)

**Variable Coefficient Std. error t-statistic Prob.** DL(GDP)t-1 0.22 0.077 2.816 0.010\*\*\* DL(INV) 0.41 0.052 7.776 0.000\*\*\* DL(INV)t-1 0.44 0.056 7.756 0.000\*\*\* DL(LE) 0.61 4.044 0.150 0.882 DL(LE)t-1 16.45 4.658 3.532 0.002\*\*\* DL(ASF) 0.39 0.437 0.888 0.384 DL(ASF)t-1 2.76 0.487 5.661 0.000\*\*\* DL(ASW) 3.18 0.940 3.379 0.003\*\*\* DL(ASW)t-1 3.91 0.909 4.302 0.000\*\*\* DL(ELC) 0.11 0.139 0.821 0.420 DL(TOP) 0.03 0.068 0.451 0.656 DL(TOP)t-1 0.49 0.072 6.819 0.000\*\*\* DUM3 0.01 0.026 0.377 0.710 ECt-1 0.53 0.048 10.931 0.000\*\*\* R-squared = 0.86; adjusted R-squared = 0.80; SER = 0.085; SSR = 0.229; LL = 56.70; AIC = 1.856;

(ASW) + 1.15 L(ELC) + 0.04 L(CO2P) + 0.80\*

*DOI: http://dx.doi.org/10.5772/intechopen.90752*

**Table 7.**

*\*\*\*, \*\*, and \* indicate significance at 1%, 5%, and 10% level, respectively.*

*ARDL long-run form. Case 2: restricted constant and no trend.*

ARDL are stable.

*ARDL short-run estimates.*

**Table 8.**

**77**

SC = 1.300; HQ = 1.648; DW = 2.15 *Note: Case 2—restricted constant and no trend;* \*\*\* *at 1% level.*
