Economic growth and human development – what do time series data say for Sudan?

Abstract

This study examines United Nations Development Programme (UNDP)'s 1996 hypothesis regarding the existence of a two-way relationship between economic growth (EG) and human development (HD) using Sudanese time series data (1960–2012). The hypothesis suggests that there is a feedback effect between HD and EG. This implies that improving HD could enhance EG opportunities and vice versa. In this study, first we analyse the long-run cointegration relationships using the Autoregressive Distributed Lag (ARDL) approach. Then, to identify the short-run dynamic relationship between HD and EG, we employ the error correction model derived from the ARDL. The results show that in the long run, HD is positively related to EG through education and employment performance channels. At the same time, EG could positively improve the opportunities for the education of households and involvement in economic activities. The short-run dynamic estimation confirms that there is a bidirectional relationship between EG and HD. The overall findings support UNDP's 1996 hypothesis concerning the existence of a two-way relationship between EG and HD. The results also confirm that improving humanity in Sudan promotes not only HD but also overall economic development.

Keywords:

• economic growth
• education
• health

JEL Classification:

• O4
• I1
• I2

Notes

1. The choice of the Lucas model instead of the Romer model is because Van Leeuwen (2007) stated that (‘Because the Lucas (1988) model is based on constant marginal returns to human capital accumulation, it is unlikely that Lucasian growth can last indefinitely. As Romer (1990) based his model on the technological frontier country (the USA), it might be possible that endogenous growth moves from Lucasian to Romerian growth when a country approaches the technological frontier’ (Phillips & Perron, 1988). Thus, since Sudan is in the primary stage of development (not reached the stage of ‘technological frontier’), we rely on the Lucas model.

2. The total physical capital (K) in an economy at any given time is referred to as capital stock. For this study, the data on real capital stock are derived from real capital formation at 2000 constant basic price using this formula ∑ (1−Kt). Where Kt is the capital stock at period T, d is the rate of depreciation, Ij is the total investment at period J and Pj is the price level at period J. IJ/PJ is the real value of the investment; in this case, it is replaced by the value of real fixed capital formation. Sudan does not provide data on physical capital stock but data on real capital formation (investment) are reported every year. For the purpose of this study, we compute real capital stock () from real capital formation by using the above formula. In the absence of specific micro surveys or information regarding the various tax legislations, the depreciation rate has been set at 10%, a choice that is in line with other studies, such as that of Bisat, El-Erian, and Helbling (1997) or Abu-Qarn and Abu-Bader (2007).

3. For why using life expectancy as an indicator for health, see Arora (2001)

4. According to the WBDI database, gross enrolment ratio (all levels combined) is the number of total students enrolled in primary, secondary and tertiary levels of education, regardless of age, as a percentage to the population of official school age for the three levels. The gross enrolment ratio can be greater than 100% because of grade repetition and entry at ages younger or older than the typical age at that grade level.

5. For more advantages on the ARDL see Pesaran and Shin (1999): Oztutk and Acaravic (2011).

6. One of the basic requirements of the ARDL is that the order of the integration between the variables must not exceed one. (i.e. No variable I(2)). (See Ozturk & Acaravic, 2011; Shahbaz et al., 2011).

7. We employ these three unit root tests because the ADF and PP tests might suffer from limited power against near unit root alternatives in finite samples (Maddala & Kim, 1998).

8. The results of when each of the LIF and UMR considered as dependent variable which is not reported here show the positive impact of the GRS on these two health measurements.

9. We employ Wald test for causality relationship instead of the traditional causality tests because the latter suffer from methodological deficiencies (see Odhiambo 2009).