INVESTMENT AND ECONOMIC GROWTH IN NIGERIA: EVIDENCE FROM VECTOR ERROR CORRECTION MODEL. Tajudeen EGBETUNDE and Isaac Olugbenga FADEYIBI

Transcription

1 Joural of Sustaiable Developmet i Africa (Volume 17, No.3, 2015) ISSN: Clario Uiversity of Pesylvaia, Clario, Pesylvaia INVESTMENT AND ECONOMIC GROWTH IN NIGERIA: EVIDENCE FROM VECTOR ERROR CORRECTION MODEL Tajudee EGBETUNDE ad Isaac Olugbega FADEYIBI Departmet of Ecoomics & Departmet of Busiess Admiistratio, Foutai Uiversity, Osogbo, Nigeria ABSTRACT The paper ivestigates the ivestmet growth exus i Nigeria, for the period Usig the Vector Error Correctio Model (VECM), the study fids that ivestmet is coitegrated with ecoomic growth i the coutry; that is, there is a log ru relatioship betwee ivestmet ad ecoomic growth i Nigeria. The results further show that ivestmet Grager causes ecoomic growth i Nigeria. The paper argues that there is eed for the govermet to ivest heavily through appropriate mechaism, strog istitutios ad macroecoomic policies i order to result to ecoomic progress ad sustaiable developmet i the coutry. Keywords: Ivestmet, Ecoomic Growth, Log Ru, Sustaiable Developmet, Vector Error Correctio Model, Grager causes, Nigeria 66

2 INTRODUCTION Ivestmet is a key factor that determies ecoomic progress i both developed ad developig ecoomies. Nigeria requires substatial ivestmet i promotig ad ehacig ecoomic activities that guaratee better livig coditios of the Nigerias. I developmet ecoomics, the debate o the ivestmet growth exus still o goig i order for developig coutries to cash up with the developed ecoomies. For ay meaigful ivestmet to take place i Nigeria, savig s habit eeds to be ecouraged. Savigs i Nigeria ecoomy was fluctuatig over the study period. Specifically, savigs habit i the coutry fell from 33.7% i 1987 to 6.9% i It dropped by 59% i 2007 ad later rose to 52% i It is importat to ivestigate this empirically for efficiet policy makigs ad formulatios. Theoretically, Nurkse (1953) argues that the so-called uderdeveloped areas, as compared with advaced, are uderdeveloped with capital i relatio to their populatio ad atural resources. Further, the emphasis was o accumulatio of physical capital to the eglect of ivestmet i educatio, health ad skills (huma capital) or techical progress. The idea was to divert a part of society s curretly available resources to icrease the stock of capital goods so as to make possible a expasio of cosumable output i the future. Nurkse did metio the vicious circle of poverty o the demad side but his solutio was a supply-side strategy of balaced growth -- "a more or less sychroised applicatio of capital to a wide rage of differet idustries" -- which he took to be a implicatio of Rosestei-Roda s theory. His view was that though "capital formatio is ot etirely a matter of capital supply... this is o doubt the more importat part of the problem." To fiace the required ivestmet a high savigs ratio (or massive foreig borrowig) would be ecessary. I the light of these assertios, it is importat to ivestigate whether the capital accumulatio i the coutry traslate to ecoomic progress ad sustaiable developmet. However, this study examies whether ivestmet spur ecoomic growth i Nigeria or ot ad covers the periods 1981 to LITERATURE REVIEW Barro (1991) examies the effect of public ivestmet ad public cosumptio expeditures o cross-coutry growth rates. After cotrollig for a umber of variables, it was foud that public ivestmet has o sigificat effect o growth rates, whereas the rate of ecoomic growth is egatively related to the share of govermet cosumptio expediture. Caig ad Fay (1993) ad Easterly ad Rebelo (1993) use pael data to ivestigate the cotributio to ecoomic growth of trasportatio etworks. A key fidig of the study is a strog relatioship betwee ecoomic growth ad public ivestmet i trasportatio ad commuicatio. Devaraja, Swaroop ad Zou (1996) preset evidece for 43 developig coutries, which idicates that the share of total govermet expediture (cosumptio plus ivestmet) has o sigificat effect o ecoomic growth. However, the authors foud a importat compositio effect for govermet expediture: that is, icreases i the share of cosumptio expediture have a sigificat positive effect o ecoomic growth, whereas icreases i the share of public ivestmet expediture have a sigificat egative effect. The egative 67

3 effect also holds for each of the major compoets of public ivestmet, icludig trasportatio ad commuicatio. This leads to the somewhat surprisig prescriptio that govermets i developig coutries would be better advised to switch public resources from ivestmet goods to curret cosumptio. Pritchett (1996) suggests aother explaatio for the Devaraja et al. (1996) fidigs the white elephat hypothesis. He argues that public ivestmet i developig coutries is ofte used for uproductive ad iappropriate projects. As a cosequece, the share of public ivestmet ca be a very poor measure of the actual icrease i ecoomically productive public capital. O the oe had, higher public ivestmet raises the atioal rate of capital accumulatio above the level chose (i a presumed ratioal fashio) by private sector agets; thus, public capital spedig may crowd out private expeditures o capital goods o a ex-ate basis as idividuals seek to re-establish a optimal iter-temporal allocatio of resources. O the other had, public capital particularly ifrastructure capital such as highways, water systems, sewers, ad airports is likely to bear a complemetary relatioship with private capital i the private productio techology. Thus, higher public ivestmet may raise the margial productivity of private capital ad thereby crowd i private ivestmet. Public ivestmet has to be a source of edogeous growth. Uder the hypothesis of balaced exogeous growth, public spedig i the log ru does ot affect ecoomic growth (Kig, Plosser, Stock ad Watsso, 1991). I a edogeousgrowth ecoomy, output follows a stochastic tred, ad permaet policy chages have log-term cosequeces for the growth of output, whereas temporary policy chages have log-term cosequeces for the level of output (Joes, 1995; Kocherlakota ad Yi, 1996; ad Evas ad Karras, 1994). I the case of edogeous growth, demad-side effects of icreased public spedig or crowdig-out effects from the way public spedig is fiaced may have log-term effects o output levels. The impact of chages i public ivestmet may vary with the level of public ivestmet. Barro (1991) specifies a edogeous-growth model, which icorporates productive public spedig (e.g., public ivestmet fiaced by lump-sum taxes) ito the productio fuctio, ad he derives a growth-maximizig spedig share. The relatioship betwee public spedig ad growth depeds o the curret spedig level; it is positive (egative) if public spedig is below (above) the growthmaximizig share. Therefore, oly whe public ivestmet is below its growth-maximizig share will additioal public ivestmet icrease growth. The govermet spedig has to take ito accout the margial effects of differet types of public spedig. The fact that public ivestmet affects output positively does ot imply that icreases i public ivestmet represet a effective growth strategy. A balaced growth strategy relies o several prerequisites. Ivestmet i high-speed broadbad telecommuicatio is required for busiesses ad istitutios (e.g., uiversities, hospitals) to fuctio efficietly ad for idividuals to commuicate. Good quality trasportatio systems are eeded to move people ad goods rapidly ad safely betwee tows ad cities. Fially, there is a commo eed for ivestmet i educatio. There is a role for govermet i promotig balaced growth. Govermet as a employer ca foster balaced growth through the decetralized provisio of public services. Aother approach would be to focus o supportig commuity ecoomic developmet (CED), icludig the provisio of capital fiacig. 68

4 Bukhari, Ali ad Saddaqat (2007) usig pael coitegratio ad causality tests, foud that both public ad private ivestmet ad public cosumptio have a log-term dyamic impact o ecoomic growth i Korea, Sigapore, ad Taiwa. The pair-wise aalysis of their study shows bidirectioal causality betwee public ivestmet ad ecoomic growth, ad the homogeeous o-causality hypothesis suggests that o-causality results are completely homogeeous i a small sample of these metioed coutries. M Amaja ad Morrissey (2008) results reveal two log ru relatios represetig the reduced form growth equatio ad the behavioural fuctio of private ivestmet. They fid that shares of private ad public ivestmet, ad imports i GDP have strog beeficial effects o per capita icome i Keya. However, aid i the form of et exteral loas is foud to have a sigificat egative impact o log ru growth. Private ivestmet relates to govermet ivestmet ad imports egatively, but positively to foreig aid. The implicatio for policy was that i order for Keya to foster ad sustai growth, closer attetio should be give to factors that promote private ivestmet. Liu, Park ad Zheg (2002) usig Grager causality aalysis, foud evidece that housig ivestmet has a stroger short ru effect o ecoomic growth tha o-housig ivestmet. They also foud that housig ivestmet has a log ru effect o ecoomic growth while ecoomic growth has a log ru effect o both housig ad o-housig ivestmet. They suggested that housig ivestmet is a importat factor for the short-term fluctuatios of ecoomic growth, with its growth stimulatig the ecoomic growth ad its slumps leadig to dowside fluctuatios. Vu (2007) usig a cross-coutry view o the impact of ICT o ecoomic growth for 50 major ICT-spedig coutries, fids that the key determiats of the variace of ICT cotributio to growth across ecoomies iclude educatio, istitutioal quality, opeess, ad Eglish fluecy. Furthermore, ICT ivestmet has a sigificat impact o ecoomic growth ot oly as traditioal ivestmet, but also as a boost to efficiecy i growth: a higher level of ICT capital stock per capita allows a ecoomy to achieve a higher growth rate for give levels of growth i labour ad capital iputs. Ju (2003) aalysig Chia s ivestmet growth exus i the cotext of the high growth experieces i East Asia ecoomies, fids that Chia has sice realized its high growth without givig rise to a icreasig proportio of ivestmet to GDP ad to arise of ICOR. He further reported that ivestmet efficiecy was largely reaped through the rural idustrializatio ad proliferatio of small firms i o-state sector. Chadra ad Sadilads (2001) show that o doubt there is a log-term positive relatioship betwee ivestmet (except govermet ivestmet where the relatioship is egative) ad GDP i Idia, but the causality is from the latter to the former ad ot vice versa. They suggest that i Idia capital accumulatio does ot cause growth i the log ru; rather growth is the cause of capital accumulatio. Haque (2013) adoptig the ew eo-classical growth model of Cobb Doglous Productio Fuctio utilizig the error correctio model (ECM), foud that the variables are statioary i first differece ad the co-itegratio tests also cofirm the existece of log term relatioship betwee the variables. The fidigs of the study cocluded that there exist a short-ru ad log-ru relatioship betwee public ad private ivestmet ad ecoomic growth i Bagladesh. 69

5 The study revealed that public ad private ivestmet impact positively o ecoomic growth i the short ad log ru process. I additio it cofirms that private ivestmet is more effective i the log ru tha public ivestmet. Accordig to the traditioal view expressed i De Log ad Summers (1991 ad 1992), fixed ivestmet i oresidetial sectors, especially i equipmet ivestmet, is the key to ecoomic growth. Blomstrom, Lipsey ad Zeja (1996), however, showed that the causal lik betwee growth ad ivestmet rus i the opposite directio. They thus suggested that the source of ecoomic growth ought to be foud somewhere else outside capital accumulatio. We (2001) fidigs do ot coform to either of these views about the cause of ecoomic growth. What he have foud, surprisigly, is that capital formatio i the residetial sector Grager-causes ecoomic growth, which i tur Gragercauses capital formatio i the busiess sector. This perhaps explais the slowdow of U.S. ecoomic growth for the 80s ad the early 90s, because residetial ivestmet as shares of GDP has bee fallig while o-residetial ivestmet as share of GDP has bee risig durig that period. I the light of the above literature, there is eed to ivestigate whether ivestmet i the Nigeria ecoomy spur ecoomic growth or ot. This will assist for effective policy makig i the coutry. THEORETICAL FRAMEWORK This study adopted AK model. A early precursor of the AK model was the Harrod-Domar model 1, which assumes that the aggregate productio fuctio has fixed techological coefficiets: Y = F(K, L) = mi {AK, BL} 1 where A ad B are the fixed coefficiets. Uder this techology, producig a uit of output requires 1/A uits of capital ad 1/B uits of labour; if either iput falls short of this miimum requiremet, there is o way to compesate by substitutig the other iput. With a fixed-coefficiet techology, there will either be surplus capital or surplus labour i the ecoomy, depedig o whether the historically give supply of capital is more or less tha (B/A) times the exogeous supply of labour. Whe AK < BL, which is the case that Harrod ad Domar emphasize, capital is the limitig factor. Firms will produce the amout Y = AK 2 ad hire the amout (1/B)Y = (1/B)AK < L of labour. Now, with a fixed savig rate, we kow that the capital stock will grow accordig to the same equatio as i the eoclassical model: K = sy δk See Harrod (1939) ad Domar (1946). 70

6 These last two equatios imply K = sak δk 4 so that the growth rate of capital will be g = K = sa δ.5 K Because output is strictly proportioal to capital, g will also be the rate of growth of output. It follows immediately that the growth rate of output is icreasig i the savig rate s. The problem with the Harrod-Domar model is that it caot accout for the sustaied growth i output per perso that has take place i the world ecoomy sice the idustrial revolutio. To see this poit, let be the rate of populatio growth. The the growth rate of output per perso is g. But if this is positive, the so is the growth rate of capital per perso K/L, sice K also grows at the rate g. Evetually, a poit will be reached where capital is o loger the limitig factor i the productio fuctio. That is, K/L will evetually exceed the limit B/A above which labour becomes the limitig factor. From the o we will istead have Y = BL, implyig that Y will grow at the same rate as L; that is, output per perso Y/L will cease to grow. METHODOLOGY AND MATERIALS The study is carried out for Nigeria for the period I this study, we measured ecoomic growth as the log of real Gross Domestic Product, GDP (deoted as Y). Gross domestic ivestmet (K) is measured as a percet of gross capital formatio to GDP 2. It is believed that other variables could have great impact o ecoomic growth ad that their omissio could bias the directio of causality betwee ivestmet ad ecoomic growth. I view of this, we icluded two cotrol variables: real iterest rate (R) ad iflatio rate (I) to avoid simultaeous bias (Gujarati, 1995) i our regressios 3. Real GDP, real iterest rate ad iflatio rate were sourced from Cetral Bak Statistical Bulleti, 2013 ad gross capital formatio as percet of GDP was sourced from World Developmet Idicator, Multivariate Co-itegratio Aalysis ad Error Correctio Modellig 2 This idicates gross domestic ivestmet i the ecoomy. 3 Asides, the icorporatio of two cotrol variables also helps to make our aalysis multivariate as agaist bivariate. This is importat because bivariate causality leads to erroeous causal ifereces (see the work of Lutkepohl, 1982; Caporale ad Pittis, 1995). 71

7 Sice the co-itegratio ad error correctio methodology i fairly commo place ad is well documeted elsewhere (Baerjee, et. al, 1993; Egle ad Grager, 1987, Johase, 1988; Johase ad Juselius, 1990) we provide oly a brief overview here. Johase (1988) multivariate co-itegratio model is based o the error correctio represetatio give by: ρ 1 X t = μ + Γ i ΔX t i + ΠX t i + ε t (6) i=1 Where X t is a (x1) colum vector of ρ variables, μ is a (x1) vector of costat terms, Γ ad Π represet coefficiet matrices, Δ is a differece operator, ad ε t N (0, Σ). The coefficiet matrix Π is kow as the impact matrix, ad it cotais iformatio about the log-ru relatioships. Johase s methodology requires the estimatio of the VAR equatio 1 ad the residuals are the used to compute two likelihood ratio (LR) test statistics that ca be used i the determiatio of the uique co-itegratig vectors of X t. The co-itegratig rak ca be tested with two statistics: the trace test ad the maximal eigevalue test. Vector Error Correctio Model (VECM) The error correctio versio pertaiig to the four variables icorporated i our study is stated below: Y t = δ 0 + δ 1i ΔY t i + δ 2i K t i + δ 3i R t i + δ 4i I t i + λecm t 1 + ε t. (7) K t = γ 0 + γ 1i ΔK t i + γ 2i Y t i + γ 3i R t i + γ 4i I t i +λecm t 1 + ε t (8) Where ECM t 1 is the error correctio term ad ε t is the mutually ucorrelated white oise residual. The coefficiet of the ECM variable cotais iformatio about whether the past values of variables affect the curret values of the variables uder study. The size ad statistical sigificace of the coefficiet of the error correctio term i each ECM model measures the tedecies of each variable to retur to the equilibrium. A sigificat coefficiet implies that past equilibrium errors play a role i determiig the curret outcomes. The short ru dyamics are captured through the idividual coefficiets of the differece terms (Akilo ad Egbetude, 2010). Ivestmet (K) does ot Grager cause ecoomic growth (Y) if all δ 2i = 0, ad ecoomic growth (Y) does ot Grager cause ivestmet (K) if all γ 2i = 0. These hypotheses ca be tested usig stadard F statistics (Mehra, 1994) 4. DISCUSSION OF RESULTS I order to avoid spurious results, we tested the variables for a uit root usig a Augmeted Dickey-Fullertest, ADF (Dickey-Fuller, 1981) ad Phillips-Perro uit root tests as show i table 1 below. 4 All the same, as poited out by Choudry (1995), Grager Causality ca still exist as log as λ is sigificatly differet from zero. 72

8 Table 1: Uiit Root Tests Augmeted Dickey-Fuller Phillips-Perro Order of Series Level First Differece Level First Differece Itegratio Y *** *** I[1] K *** *** I[0] R *** *** I[1] I *** *** I[0] ***, **, * idicate sigificace level at 1%, 5% ad 10% respectively. Figures i parethesis are t-statistic. The results of the statioarity tests at level show that all the variables are o-statioary at level. Havig foud that the variables are ot statioary at level, the ext step is to differece the variables oce i order to perform statioarity tests o differece variables. The results of the statioarity tests o differeced variables cofirmed statioarity. Havig cofirmed that all variables icluded i the causality test are itegrated of order oe, the ext step is to test for the existece of a coitegratio relatioship amog the variable series usig the Johase-Juselius approach described i the methodology. The coitegratio test results are reported i Table 2. Table 2: Coitegratio Test Series: Y K R I Max-Eige Stat. Trace Stat Critical Prob.** Hpothesized No. of CE(s) Value Noe * At most At most At most 3 * ***, **, * idicate sigificace level at 1%, 5% ad 10% respectively. Figures i parethesis are t-statistic. The results idicate the existece of coitegratio relatioship amog the variables. The maximum eigevalue ad trace statistics reject the ull hypothesis of o coitegratio at 5 percet level.. Accordig to N Zue (2006), whe coitegratio exists, the Egle-Grager Theorem establishes the ecompassig power of the ECM over other forms of dyamic specificatio. Results of the ECMs are preseted i Table 3 below. Sice the error correctio represetatio ca be used to test for Grager causality, we estimated both equatios 7 ad 8. 73

9 Table 3: VECM Results Growth Model Ivestmet Model ECM(-1) *** ECM(-1) *** [-4.522] [-3.647] D(Y(-2)) 0.606*** D(K(-2)) *** [ 3.073] [-3.884] D(K(-1)) D(Y(-1)) 0.005*** [-1.390] [ 3.233] D(K(-2)) D(Y(-2)) [-1.448] [ 0.558] D(K(-3)) D(R(-1)) [-1.596] [ 0.142] D(R(-1)) *** D(R(-2)) [-3.127] [-0.925] D(R(-2)) *** D(I(-1)) 0.043* [-4.110] [ 1.930] D(I(-1)) 0.788*** D(I(-2)) [ 4.570] [-0.508] D(I(-2)) 0.728*** C [ 4.590] [ 0.236] D(I(-3)) 0.458*** R-squared [ 3.407] Adj. R-squared C F-statistic R-squared Adj. R-squared F-statistic F-stat {Y K} [ 0.972] F-stat {K Y} 35.07*** ***, **, * idicate sigificace level at 1%, 5% ad 10% respectively. Figures i parethesis are t-statistic. F-stat {Y K} deotes F-statistics that ecoomic growth Grager causes ivestmet.. F-stat {K Y} deotes F-statistics that ivestmet Grager causes ecoomic growth.. 74

10 I geeral, the results show that the error correctio terms i both equatios are well defied, that is, their associated coefficiets are egative ad statistically sigificat at 1% (see table 3). The coefficiet was i growth equatio ad i ivestmet equatio of the models i Nigeria. This idicates a feedback of approximately 73 percet (for growth equatio) of the previous year s disequilibrium ad a feedback of approximately 47 percet (for the ivestmet equatio) of the previous year s disequilibrium i Nigeria. The strog sigificace of the coefficiet o ECM supports the coclusio of coitegratio. This idicates that ivestmet serves as a catalyst to ecoomic growth i Nigeria, ad also icome is a egie of ivestmet which i tur trasforms the level of ecoomic progress i the coutry. Moreover, if the rate of ivestmet i the coutry is sustaied ad ehaced, the coutry will witess sustaiable developmet which i tur improve stadard of livig of people that residet i the coutry. The short ru dyamics are captured by the idividual parameters except that of the ECM term. The results (i growth model) show that ivestmet is isigificatly related to ecoomic growth i Nigeria. This could be as a result of factors 5 that are iimical to ivestmet s motivatio which further worseig the livig stadard i the coutry. The results (i growth model) further reveal that iterest rate (lagged by oe to two) is egative ad sigificat impacted o ecoomic growth i the ecoomy. 6 Also, iflatio (lagged by oe to three) has positive ad sigificat effect o ecoomic developmet i Nigeria. This suggests that moderate iflatio boost ecoomic activities i the coutry. 7 I the ivestmet model, it was evideced that ecoomic growth (lagged by oe) has a positive ad sigificat effect o ivestmet i Nigeria. This suggests that advacemet i ecoomic growth goes alog with adequate ivestmet i the coutry. More importatly, this study suggests that ecoomic growth is oe of the key factors that drive ad ehace ivestmet activities i Nigeria. It was further revealed i the ivestmet model that iflatio rate (lagged by oe) has positive ad sigificat ifluece o ivestmet i the coutry. This implies that moderate iflatio rate ecourages ivestmet activities i the coutry. As evideced from the stadard F-Statistics 8 reported i Table 3, there is uidirectioal causality ruig from ivestmet to ecoomic growth i Nigeria. The results show that ivestmet Grager causes ecoomic growth i the coutry. This implies that ivestmet promotes ecoomic growth i Nigeria. This result supported the view of orthodox developmet ecoomists propositio o the requiremet to attai developmet i developig coutries. Furthermore, Nigeria ecoomy should ehace ivestmet activities through appropriate mechaisms which i tur result to sustaiable developmet i the coutry. 5 The factors iclude iterrupted power supply, poor road etwork facilities, iadequate skilled mapower, amog other. 6 This implies that high iterest rate slow dow developmet strategies i the coutry. This high iterest rate discourages ecoomic agets to embark o programmes that may serve as catalyst to developmet of Nigeria ecoomy. 7 This view is supported by scholars i developmet ecoomics. It was postulated by scholars i this field that high iflatio poited to rise i volume of moey ad ivestors use the opportuity to ivest heavily for the progress of a ecoomy. 8 This is the results obtaied from multivariate VECM Grager Causality/Block Exogeeity Wald Tests. 75

11 F-statistic i both models is sigificat at 5%. This implies the fitess of the growth ad ivestmet models that this study estimated. Ad the adjusted R-squared is ot too high which trash out the issue of multicolliearity i the models. CONCLUDING REMARKS This paper ivestigates the impact of ivestmet o ecoomic growth i Nigeria usig multivariate Grager causality test withi the cotext of VECM framework. The paper also examies the log ru coitegratig relatioship amog the series. The results reveal that all the variables are coitegrated i.e. they are related i the log ru. We also cofirm this coitegratio tests through ECM test withi the VECM framework, the fidig shows that a log ru relatioship exist betwee ivestmet ad ecoomic growth i the coutry. This result corroborates the log ru relatioship amog the series. Moreover, the results further reveal that ivestmet Grager causes ecoomic growth i Nigeria. Therefore, govermet i the coutry should put more efforts i accumulatig capital to further ehace improvemet ad efficiecy of the ivestmet activities which i tur accelerate sustaiable ecoomic developmet i the coutry. Also the coutry should adopt policies that will fast track sustaiable developmet through heavy ivestmet, ad the govermet should esure the provisio of adequate facilities that promote ad ecourage heavy ivestmet. This otio supported the view of Leibestei (1957) who suggested that the oly way of escapig the trap of uderdevelopmet is to raise per capita icome to a level where growth becomes self-sustaiig. This was liked with the idea of big push or critical miimum effort i developmet ecoomics. I geeral, the evidece from the paper suggests that policy makers i the coutry should ecourage ivestors (both local ad iteratioal) through appropriate mix of strog istitutios, legal ad regulatory policies i order to provide eablig eviromet for ay busiess to thrive. REFERENCES Akilo, A.E. ad Egbetude, T. (2010). Fiacial developmet ad ecoomic growth: the experiece of 10 sub-sahara Africa coutries revisited. The Review of Fiace ad Bakig, Volume 02, Issue1, Baerjee, A., Dolado, J.J., Galbraith, J.W., ad Hedry, D.F. (1993). Coitegratio, error correctio, ad ecoometric aalysis of o-statioary data. Oxford: Oxford Uiversity Press. Barro, R.J. (1991). Ecoomic growth i a cross-sectio of coutries. Quarterly Joural of Ecoomics, Vol. 106, Blomstrom, M., Lipsey, R.E. ad Zeja, M. (1996). Is fixed ivestmet the key to ecoomic growth?. Quarterly Joural of Ecoomics, February, Bukhari, S.A.H.A.S., Ali, L. ad Saddaqat, M. (2007). Public Ivestmet ad Ecoomic Growth i the Three Little Dragos: Evidece from Heterogeeous Dyamic Pael Data Caporale, G.M. ad Pittis, N. (1995). Causality iferece i bivariate ad trivariate systems: some more results. Discussio paper No. DP15 95, Cetre for Ecoomic Forecastig, Lodo Busiess School. De Log, J.B. ad Summers, L. (1991). Equipmet ivestmet ad ecoomic growth. Quarterly Joural of Ecoomics, Vol. CVI, De Log, J.B. ad Summers, L. (1992). Equipmet ivestmet ad ecoomic growth: How strog is the exus? Brookigs Papers o Ecoomic Activity, Devaraja, S., Swaroop, V. ad Zou. H. (1996). The compositio of public expediture ad ecoomic growth. Joural of Moetary Ecoomics, Vol. 37,

12 Domar, E. D. (1946). Capital Expasio, Rate of Growth, ad Employmet. Ecoometrica, Vol. 14, Caig, D., ad Fay. M. (1993). The effect of trasportatio etworks o ecoomic growth. Columbia Uiversity Departmet of Ecoomics Discussio Paper. Chadra, R. ad Sadilads, R.J. (2001). Does ivestmet cause growth? A test of a edogeous demad-drive theory of growth applied to Idia Uiversity of Strathclyde, UK, pp1-21. Easterly, W., ad Rebelo. S. (1993). Fiscal policy ad ecoomic growth. Joural of Moetary Ecoomics, Vol. 32(3), Egle, R.F. ad Grager, C.W.J. (1987). Coitegratio ad error correctio: represetatio, estimatio, ad testig. Ecoometrica, Vol. 55, Evas, P. ad Karras, G. (1994). Are govermet activities productive? Evidece from a pael of U.S. states. Review of Ecoomics ad Statistics, Vol. 76, Gujarati, D. (1995). Basic ecoometrics. New York: Mcgraw-Hill. Johase, S. (1988). Statistical aalysis of coitegratig vectors. Joural of Ecoomic Dyamics ad Cotrol, Vol. 12, Johase, S. ad Juselius, K. (1990). Maximum likelihood estimatio ad iferece o coitegratio with applicatio to demad for moey. Oxford Bulleti of Ecoomics ad Statistics, Vol. 52, Joes, C.I. (1995). Time series tests of edogeous growth models. Quarterly Joural of Ecoomics, Vol. 110(2), Ju, Z. (2003). Ivestmet, ivestmet efficiecy, ad ecoomic growth i Chia. Joural of Asia Ecoomics, Vol. 14, , Published by Elsevier Ic. Haque, S.T. (2013). Effect of Public ad Private Ivestmet o Ecoomic Growth i Bagladesh: A ecoometric Aalysis. Research Study Series, No FDRS 05/2013, Harrod, R. F. (1939). A Essay i Dyamic Theory. Ecoomic Joural, Vol. 49, Kig, R.G., Plosser, C.I., Stock, J.H. ad Watso. M.W. (1991). Stochastic treds ad ecoomic fluctuatios. America Ecoomic Review, Vol. 81(4), Kocherlakota, N. R., ad Yi, K. (1996). A simple time series test of edogeous vs. exogeous growth models: a applicatio to the Uited States. Review of Ecoomics ad Statistics, Vol. 78, Leibestei, H. (1957). Ecoomic backwardess ad ecoomic growth. New York, Wiley. Liu, H., Park, Y.W. ad Zheg, S. (2002). The Iteractio betwee Housig Ivestmet ad Ecoomic Growth i Chia. Iteratioal Real Estate Review, Vol. 5, No. 1, Lutkepohl, H. (1982). No-causality due to omitted variables. Joural of Ecoometrics, Vol. 19, M Amaja, D. ad Morrissey, O. (2008). Foreig Aid, Ivestmet, ad Ecoomic Growth i Keya: A Time Series Approach. CREDIT Research Paper, No. 06/05. Cetre for Research i Ecoomic Developmet ad Iteratioal Trade, Uiversity of Nottigham Mehra, P.Y. (1994). Wage growth ad the iflatio process: a empirical approach, i B.B. Rao (ed.), Coitegratio for the Applied Ecoomist, , New York: St Marti s Press. Nurkse, R. (1953). Problems of capital formatio i uderdeveloped coutries. Oxford Uiversity Press, New York. 77

13 N zue, F.F. (2006). Stock market developmet ad ecoomic growth: evidece from Cote D Ivoire. Africa Developmet Review, Vol. 18(1), Prichett, L. (1996). Mid Your p s ad q s: the cost of public ivestmet is ot the value of public capital. World Bak Workig Paper, Vu, K. (2007). Measurig the impact of ICT ivestmets o ecoomic growth. Program o Techology ad Ecoomic Policy, Harvard Keedy School of Govermet, We, Y. (2001). Residetial Ivestmet ad Ecoomic Growth. Aals of ecoomics ad fiace, Vol. 2, ABOUT AUTHORS: Tajudee EGBETUNDE (Ph.D., Ecoomics) is a seior lecturer i the Departmet of Ecoomics, Foutai Uiversity, Osogbo, Nigeria. Isaac Olugbega FADEYIBI is a lecturer I i the Departmet of Busiess Admiistratio, Foutai Uiversity, Osogbo, Nigeria. 78