Production linkages and dynamic fiscal employment effects of the extractive industries: input-output and nonlinear ARDL analyses of Azerbaijani economy

Abstract

In this paper, we address the production linkages and employment effects of the petroleum sector on the rest of the Azerbaijani economy. The availability of the input-output tables for the years 2006, 2008, and 2009 enables the assessment of the changes with regard to the multiplier effects of the extractive industries over the first 3 years of the oil boom. We find that despite advanced infrastructure, well-developed petrochemical complex, and local content policies, the degree of integration of the international oil and gas business into the domestic economy is rather weak. In addition, both production and job creation multipliers slightly decreased after 3 years of exponential growth rates of oil production. The assessment of the production multipliers indicates that additional investments in processing, construction, and network industries have the highest production linkages. Concerning employment multipliers agriculture, education, health care, and public sector have the greatest job creation effects. To assess the fiscal employment effects of the oil revenues, which cannot be captured over the static input-output analysis, we employ the cointegrating nonlinear autoregressive distributed lag model. The model reveals a sustainable job creation effect of oil revenues in the case of Azerbaijan.

Appendices

Appendix

Table 4. Simulation of the 1 Mio. USD increase of the oil sector output (final demand) based on IO tables of 2009

Table 5. Simulation of the 1 Mio. USD increase of the non-oil sector output (final demand) based on IO tables of 2009

Table 6. Direct linkage effects of extractive industries on the other sectors 2006

Table 7. Direct linkage effects of processing industries on the other sectors of economy 2006

Table 8. Job creation effect of USD 1 Mio. Output growth on the economy-wide employment

Table 9. Leontief multiplier by sector for 2006 and 2009

Appendix 7. IO analysis–based assessment of the employment effects

The aggregate output of all sectors in the economy is the sum of the intermediate, AX, and the final output, Y:

$$ X= AX+Y $$

(A.1)

whereby, A is the direct expenses coefficient matrix which is also called technological matrix for representing production technology. X is the total output of goods and services, and Y is the last product vector. Equation (A.1) could be rewritten as follows in Eq. (A.2), whereby E is the respective identity matrix

$$ X={\left(E-A\right)}^{-1}Y $$

(A.2)

We start the presentation of the labor balance model by introducing the direct labor coefficient t_i, which is calculated as the ratio of the total employment and total output in the individual sectors. This corresponds to Eq. (A.3):

$$ {t}_j=\frac{L_j}{X_j},\kern0.5em i=\overline{1,n} $$

(A.3)

whereby X_i is the total product of the j-th sector and L_j - is the labor force which is employed in the j-th sector for the production of the j-th product. As already mentioned, t_i captures only the direct job creation effect. If we denote the full job creation coefficient as T_j, then

$$ {T}_j=\sum \limits_{i=1}^n{a}_{ij}{T}_i+{t}_{j,}=\overline{1,n} $$

(A.4)

whereby a_ij is the direct material expense coefficient and T_j is full the labor coefficient. If we include direct labor coefficient t = (t₁, t₂, …, t_n) and the full labor coefficient T = (T₁, …. T_n) as a line vector, then Eq. (A.4) can be written in the matrix form (A.5):

$$ T=T\ A+t $$

(A.5)

The multiplication of the LHS of Eq. (A.5) by the identity matrix, E, yields:

$$ T=t{\left(E-A\right)}^{-1} $$

(A.6)

(E − A)⁻¹ matrix is the well-known full material expense coefficients matrix. If we denote (E − A)⁻¹ as B, then Eq. (A.6) can be rewritten as:

$$ T= tB $$

(A.7)

Equation (A.3) yields:

$$ {L}_j={t}_j{X}_j $$

If we assume that there are n sectors in the economy, then the whole labor force which is employed in the economy, L, can be calculated in accordance with Eq. (A.8):

$$ L=\sum \limits_{j=1}^n{t}_j{X}_j= tX $$

(A.8)

The multiplication of the both sides of (A.7) by the final product vector, Y yields

$$ TY= tBY $$

(A.9)

From (A.2), we know that BY = X. Hence, Eq. (A.9) can also be expressed as follows:

$$ T\ Y=t\ X $$

(A.10)

Because, the RHS of the last equation equals the RHS of Eq. (A.8), which shows the total number of the employed people, we can rewrite Eq. (A.10) as follows:

$$ L= TY $$

(A.11)

This equation shows that the full labor coefficient T and determines the number of the additional employees to produce the final product. Based on Eq. (A.11), the job creation, ∆L, can be calculated as ΔL = tΔX = tB∆Y = T∆Y.³ ΔL stands for the direct and indirect job creation effects caused by the change in the amount of the final good, ΔY.