Currency equivalents

B. The Volatility of Fiscal Variables

2.10 The high dependence on oil brings high volatility to fiscal variables, especially revenues. Macroeconomic volatility reflects the effect of external shocks hitting the economy. Terms of trade fluctuations appear to be a major force behind fiscal volatility (World Bank 2000), and this is particularly true with oil prices for Algeria. Non-hydrocarbon GDP has been less volatile than total GDP. More important, most revenue variables are more volatile than expenditure variables (Table 2.3). This is so because oil price disturbances have an immediate impact on public revenues: Total and hydrocarbon revenue have been more volatile than total public expenditures. Furthermore, current and capital expenditures exhibit lower volatility than GDP, which is an unexpected finding as more variation in capital expenditure, particularly during a long adjustment period, would have been a more conventional result. Predictable, however, is the lowest volatility found in wages and salaries (1.0), which already indicates their inertial character.

Table 2.3 Volatility of Fiscal Variables, 1990–2005 (in percent of GDP unless noted)
	Mean	Standard Deviation
Total revenue	32.7	4.1
Hydrocarbon	21.1	4.9
Nonhydrocarbon	11.6	1.2
Tax revenue	10.5	1.2
o/w VAT	4.9	0.7
Total expenditure	30.6	2.8
Current expenditure	21.9	2.2
Interest	2.8	1.0
Wages and salaries	8.3	1.0
Capital expenditure	8.1	1.5
Net lending	0.6	1.5
Primary balance	4.8	5.5
Budget balance	2.7	4.9
NH balance / NH GDP	-27.5	4.3
Domestic financing	-2.8	5.0
External financing	0.7	3.4
GDP (real growth)	2.7	2.6
Non-hydrocarbon GDP (real growth)	2.3	3.1

Source: Bank staff calculations.

Note: NH=Non-hydrocarbon balance = NH revenues/NH GDP.

11. 
    Bivariate correlation analysis shows a significant association between practically all fiscal variables and per capita growth.¹⁸ Simple correlations reported in Table 2.4 present findings that are consistent with previous literature on the relationship between fiscal variables and growth (Gupta et al. 2005). For example, budget surpluses (and primary balances), and revenues (especially hydrocarbon) are positively associated to per capita growth (and most of them statistically significant). In a similar vein, financing needs, either domestic or external, and the Non-hydrocarbon balance are negatively associated to per capita growth.
    

Table 2.4 Bivariate Correlations, 1990–2005
(Variables expressed as percent of GDP, unless otherwise specified)
	Per capita real GDP growth
Total revenue	0.46	*
Hydrocarbon revenue	0.47	*
Nonhydrocarbon revenue	-0.19
Tax revenue	-0.35
o/w VAT	-0.24
Total expenditures	-0.32
Current expenditures	-0.27
Interest	-0.04
Wages and salaries	-0.62	***
Capital expenditures	0.40
Net lending	-0.64	***
Primary balance	0.49	**
Budget balance	0.41
Non-oil balance / NH GDP	-0.07
Domestic financing	-0.36
External financing	-0.37
Source: World Bank calculations
* Significant at 10%; ** significant at 5%; ***significant at 1%.
Note: Bilateral correlations using annual data from 1990–2005.

2.12 Even more important, bivariate correlation analysis confirms the procyclical character of capital expenditures and the importance for Algeria to preserve the Hydrocarbon Stabilization Fund created in 2000 (see Box 2.1 below). The composition of public expenditure matters for growth: higher capital outlays are associated with higher growth, while lower current expenditure is associated with more favorable economic conditions. An IMF country report also finds that government capital spending induces an increase in real nonhydrocarbon GDP, while current expenditures do not (IMF 2005b). For its part, correlation analysis reveals the negative correlation of wages and salaries with growth, which is also indicative of their inertial trends^.19.