The use of econometric methods in the constructing of mathematical systems for detecting «Dutch disease» in economy. Evidence from Nigeria

At the present stage of economic development, many countries faced the problem associated with the movement of factors of production from less developing sectors of the economy in a rapidly developing industries. Typically, such movement includes the capital, as investments in the fast-growing sector become immediately profitable. Furthermore, human resources in such sectors are better compensated as a result of increased demand . Reallocation of capital and labor can have a negative impact on those industries that have lost these resources, causing them to decline. As the cost of production factors exacerbated by the effect of the resources movement, those sectors competing for affected resources tend to lose their competitive edge. To sum up, Dutch disease affects the economy of a country in two ways: first by causing a resource movement effect from one sector of the economy to another, and second, by inducing a spending effect that appreciates the real exchange rate of the country’s currency¹. Symptoms listed above make it possible to ascertain the presence of "Dutch Disease" in a particular country.

In Africa, there is a looming threat of Dutch disease due to the substantial natural resource wealth of many of its regions. Nigeria began extracting oil in the late 1950s and since then the oil industry has dominated the economy and given it much needed boosts. However, price shocks in oil and recessions have severely damaged its economy, leaving it at the mercy of the market.

Literature review

There is a growing literature dealing with different aspects of the Dutch Disease and resource curse. Since the focus of this article is on the existing of «Dutch Disease» in Nigeria and an examination of economic indicators which affect GDP more , this section reviews some empirical and theoretical studies on this topic which were used in analysis.

Theoretical study of Jason Gould and Katen N. Kapadia High explains the effects of increasing in oil production which leads to ignoring traditionally strong sectors. In the other study, Bayoumi and Mhleisen (2006) estimate the influence of changes in energy and non-energy commodity prices on real exchange rate. The work of Forsyth and Nicholas (2008) was also dedicated to examining whether economy of the country suffer from «Dutch disease». Their general result is that non-oil manufacturing was spared the perverse effects of oil price increases despite the appreciation of the nominal and real exchange rate.

On the basis of mentioned studies I composed system of equations, which helps to test economy of Nigeria for the symptoms of «Dutch disease».

Empirical evidence

To confirm the hypothesis of existing "Dutch disease" in economy of Nigeria I took yearly data from 1995 to 20152. The analysis in this article is based on the World Bank's official data. The following indicators were analyzed:

• 1.    Gross domestic product (Billion US dollars);

• 2.    Total export of goods (Billion US dollars) ;

• 3.    Fuel export (Billion US dollars);

• 4.    Consumer price index;

• 5.    Exchange rate of Nigerian Naira against US dollar

• 6.    Oil prices

• 7.    Foreign direct investment inflow.

Using this indicators I composed following equation:

(            Y = C-l + C2Xi + C3x4 + C4x 6 ;

[% 4 = C5 + C₆^Y + C7x 2 + C8^x 5 ^{+ C}9^x6 ⁺ Cl 0 ^x 7 ⁺ Cl l x 3 .

Where Y – GDP, X 1 – Total export of goods; X 2 – Fuel export, X 3 – Real interest rate , X 4 – CPI, X 5 – exchange relation US$/Nigerian Naira, X 6 – oil prices, X 7 – FDI .

According to model above, dependent variables are GDP expressed in billions US dollars and Consumer Price Index. Total export of goods, fuel export, real interest rate, exchange relation US$/Nigerian Naira, oil prices and foreign direct investment are considered as instrumental variables

With the help of Eviews, using Two- Stage Least Squares Method I obtain following results:

	Coefficient	Sid. Error	t-Statistic	ProD
C(1)	-59.33020	25.75410	-2 303719	0 0281
0(2)	3.560144	1.550658	2 295893	0 0286
C(3)	2 954952	0.460034	6.423332	0.0000
C(4)	-2.776927	1 819132	-1 526470	0 1370
C(5)	-8 697240	9.150549	-0 950461	0 3492
C(6)	0.142776	0.035977	3.968519	0.0004
C(7)	-0 267654	0 083657	-3 199435	0 0032
C(8)	0 290197	0.089809	3.231284	0 0029
C(9)	0.322457	0.125900	2.561203	0.0155
C(10)	14 36297	5.741797	2 501475	0.0179
C(11)	0015436	0 084556	0.182549	0 8563
Determinant residual covariance		28879.13

Looking at model probability of mistakes of coefficient C(5) and C(11) is high and it exceeds allowable 15%. In this case I decided to exclude coefficient

C(11) from the model to make it more significant. C(5) is constant that is why we don’t take its value into account and don’t exclude from the model.

	Coefficient	Std. Error	t-Statistic	Prob.
0(1}	-58.03160	26.03623	-2 228879	0.0330
0(2)	3.640109	1.567715	2.321920	0.0268
0(3}	2.897192	0.467408	6.198419	0.0000
0(4}	-2.804865	1.837495	-1.526461	0.1367
0(5)	-9.223577	7.435094	-1.240546	0.2238
0(6)	0.141646	0.031673	4.472171	0.0001
0(7)	-0.266441	0.080261	-3.319680	0.0023
0(0)	0.296166	0.070901	4.177164	0.0002
0(9}	0.319902	0.123813	2.583749	0.0145
0(10)	14.53747	5.162388	2.816036	0.0083
Determinant residual covariance		29780.66

Equation: Y= C(1) * C(2)*X1 +C(3)*X4 + C(4)*X6

Instruments: X1 X2 X5 X6 X7 C

Observations: 21

R-squared Adjusted R-squared S.E. of regression Durbin-Watson stat

0.948815  Mean dependent var      194.3020 0.939783  S.D. dependentvar       187.6431 46.04620  Sum squared resid       36044.29 1.061322

Equation: X4 = C(5) + C(6)*Y * C(7)*X2 +C(8)”X5 +C(9)*X6 * C(10)*X7

Instruments: X1 X2X5X6X7 C

Observations: 21

R-squared Adjusted R-squared S.E. of regression Durbin-Watson stat

0.990391  Mean dependent var      69.61427 0.987189  S.D. dependentvar       44.38420 5.023751  Sum squared resid       378.5712 2.378715

High value of coefficient of determination R2=0.94 shows that 94% of total deviation of GDP is explained by the variation of total export of goods, consumer price index and oil prices. This result proves all chosen indicators to be significant. The other coefficient of determination of second equation is also high3. It illustrates that fuel export, real interest rate, oil prices, FDI inflow significantly influence on CPI. According to the model above we can see that probability of mistakes of all coefficients is acceptable.

Then, to check whether autocorrelation is of the residuals of the model, we should pay attention on value of Durbin- Watson statistic. With the help of tables of Durbin- Watson statistic , we can find that for n = 21 and k = 5 values d L = 0,83 and d U = 1,96. As we see , DW 1 = 1,06 is on the interval (0,83 : 1,96 ) in uncertainty zone and because of this we can’t argue whether there is autocorrelation or not. The second value DW 2 = 2, 38 is in uncertainty zone as well (4 – 1,96 = 2, 04), it is also impossible to conclude about the presence of autocorrelation.

Conclusion

Analyzing the econometric model we can see that total export of goods has a large influence on Gross domestic product. With an increase in exports to one billion US dollars, GDP increases by 3.6 billion dollars. In addition, exchange-traded oil prices also have a significant impact on GDP. With an increasing of oil prices on the US dollar, the GDP of Nigeria loses 2.8 billion US dollars. Keeping at view that oil exports exceed 70% of the total export of goods in the structure of exports of Nigeria, we can conclude that this is the dominant sector of the economy, while other industries remain outside the attention of the state and stay underdeveloped. This factor lets us say that economy of Nigeria suffer from «Dutch disease».

However, analyzing the second equation, we can see that an increase in FDI to one billion US dollars, the CPI increases by 14 units. In this case, the task of the government is to organize measures to attract investment in other sectors of the economy. Thus, the development of other sectors will reduce the dependence of the Nigerian economy on the world market price changes.

«Экономика и социум» №2(33) 2017