Some economists argue that the composition of the budget plays a vital role in the process of economic growth ( Gupta et al., 2005 ). These authors consider that economic growth evolves faster in countries where capital expenditure predominates in total government expenditure. Let us examine the composition of public expenditure in Senegal, per three (3) year period from 2010 to 2021 ( Figure 1 ):

This graph shows that in Senegal, current expenditure takes up a greater proportion of the state budget. Over the period 2010-2017, public consumption expenditure represents, on average, 62% of the budget, while public capital expenditure is around 38%. This strong pressure on public consumption expenditure during this period is mainly linked, on average, to the rapid evolution of the payroll at 8.8% and interest on the public debt at 17.5%. During this same period, the increase in public capital expenditure was 20.58%, due in particular to the allocation of significant internal resources in favour of public capital expenditure (30.4%) and, to a lesser extent, to external resources (7.7%) according to Central Bank of West African State.

However, it should be noted that from 2017 to 2021, the gap between the share of current expenditure and that of capital expenditure widened further and is explained by:

- The deceleration of public investments financed by external resources;

- The implementation of the Economic and Social Resilience Plan (PRES) through budgetary transfers;

- The increase in the number of civil service staff and the financial impact of the budgetary commitments made by the State to the social partners in education and health, in particular the tenure of teachers and contract education professors.

In some developing countries, the State often plays the role of employer of last resort. In addition to the problem that this role can pose in terms of productivity, it means that salary expenditure tends to represent a significant share of public expenditure, which is the case in Senegal. Such a situation leaves limited resources to finance productive spending in education, infrastructure, and health, which are capable of promoting growth and human development.

The following Figure 2 provides a better understanding of the evolution of debt interest in relation to total current expenditure:

Despite the rapid evolution of current expenditure in the state budget, Senegal has been able to control the trajectory of interest on its public debt. From 2010 to 2015, public debt interest accounted for an average of 3.81% of total current expenditure and, between 2016-2020, it was around 2.45% on average. Therefore, a fundamental question arises: what is the contribution of budgetary expenditure to economic growth?

In the following subsection, we examine the contribution of public spending, private sector spending and the external sector to Senegal’s economic growth.

Determining the contribution to GDP growth amounts to calculating the share of each component of GDP in constituting the economic growth rate. Based on data from Senegal’s national accounts over the period 2010-2020, we start from the following formula:

$\text{In}t,GD{P}_t=C{g}_t+C{p}_t+I{g}_t+I{p}_t+X_t-M_t$ (1)

$\text{In}t-1,GD{P}_{t-1}=C{g}_{t-1}+C{p}_{t-1}+I{g}_{t-1}+I{p}_{t-1}+X_{t-1}-M_{t-1}$ (2)

If we differentiate between Equations (1) and (2), we obtain the following equation:

$\begin{array}{c}GD{P}_t-GD{P}_{t-1}=\left(C{g}_t-C{g}_{t-1}\right)+\left(C{p}_t-C{p}_{t-1}\right)+\left(I{g}_t-I{g}_{t-1}\right)\\ +\left(I{p}_t-I{p}_{t-1}\right)+\left(X_t-X_t{}_{-1}\right)-\left(M_t-M_{t-1}\right)\end{array}$ (3)

By dividing each member of Equation (3) by $GD{P}_{t-1}$, we then obtain Equation (4):

$\begin{array}{c}\frac{GD{P}_t-GD{P}_{t-1}}{GD{P}_{t-1}}=\frac{C{g}_t-C{g}_{t-1}}{GD{P}_{t-1}}+\frac{C{p}_t-C{p}_{t-1}}{GD{P}_{t-1}}+\frac{I{g}_t-I{g}_{t-1}}{GD{P}_{t-1}}\\ +\frac{I{p}_t-I{p}_{t-1}}{GD{P}_{t-1}}+\frac{X_t-X_{t-1}}{GD{P}_{t-1}}-\frac{M_t-M_{t-1}}{GD{P}_{t-1}}\end{array}$ (4)

However, the growth rate of GDP $\left(Gr\right)=\frac{GD{P}_t-GD{P}_{t-1}}{GD{P}_{t-1}}$ We can therefore affirm:

$\begin{array}{c}Gr=\frac{C{g}_{t-1}}{GD{P}_{t-1}}\times GCg+\frac{C{p}_t{}_{-1}}{GD{P}_{t-1}}\times GCp+\frac{I{g}_{t-1}}{GD{P}_{t-1}}\times GIg\\ +\frac{I{p}_{t-1}}{GD{P}_{t-1}}\times GIp+\frac{X_{t-1}}{GD{P}_{t-1}}\times GX-\frac{M_{t-1}}{GD{P}_{t-1}}\times GM\end{array}$ (5)

with

Gr = Economic growth rate at time t;

GCg = Growth rate of public consumption at time t;

GCp = Growth rate of private consumption at time t;

GIg = Growth rate of public investments at time t;

GIp = Growth rate of private investments at time t;

GX = Growth rate of exports at time t;

GM = Growth rate of imports at time t.

The contribution calculation is the same for each economic sector. We present the details for the government sector:

The sum of $\frac{C{g}_{t-1}}{GD{P}_{t-1}}\times GCg$ and $\frac{I{g}_{t-1}}{GD{P}_{t-1}}\times GIg$ corresponds to the contribution of the government sector to economic growth in year t.

This formula for calculating the economic growth rate (Gr) shows that the economic growth rate of a nation is a weighted average of the growth rate of the other components of aggregate demand. The weighting used here is the component of the overall demand in t − 1 compared to the GDP of the previous period. Figure 3 represents the contributions of the public, private and rest of the world sectors to Senegal’s economic growth as follows:

This graph shows us that the Senegalese economy is driven from 2000 to 2021 by the private sector, with the public sector coming in second place. In a logic of carrying out development programs and raising low growth rates, the Government uses the budget to revive the growth of economic activity and influence economic variables such as household consumption, private investments and jobs. The increase in budgetary expenditure with a strong knock-on effect on the economy, such as the creation of airports, toll highways and flyovers, is often undertaken by the Government to accelerate the level of economic activity.

However, we also note the non-competitiveness of the Senegalese economy, which results in a negative contribution from its external sector. Indeed, imports of goods and services far exceed exports of goods and services and the trade balance is permanently in deficit.

Since we are interested in short and medium-term economic growth, we will present the average economic contributions per 10-year period ( Table 1 ):

Between the two decades, there was a drop in the private sector contribution of 2.15 points and an increase in the public sector contribution of 0.95 points. This situation demonstrates the presence of a crowding-out effect (or crowding-out effect) and reveals the limits of an overly active budgetary policy. This effect refers to the contraction of household consumption expenditure and private investment by interest rate-sensitive companies following an increase in public spending financed by debt. The crowding-out effect is explained by the transmission mechanism between the money market and the market for goods and services. The increase in public spending leads to an increase in the demand for transaction money and the interest rate. Such reactions attenuate the positive effects of state fiscal policy.

Saunders (1985) considers that the economic performance indicators used are the rate of economic growth, the rate of consumer price inflation and private sector employment growth. Overall, the results provide little evidence that government size and growth have been detrimental to economic performance, particularly since 1975, although an inverse relationship existed between government size and economic growth in the sixties.

Aschauer (1989) , in his work carried out in the United States, finds significant results over the period 1945-1985. In his model on American data, he estimates, on American time series, the following model:

Aschauer (1989) , in his work carried out in the United States, finds significant results over the period 1945-1985. In his model on American data, he estimates, on American time series, the following model:

$\log \left(Production\right)={\beta }_0+{\beta }_1\log \left(Kpr\right)+{\beta }_2\log \left(Kpub\right)+{\beta }_3\log \left(Kpub\ast Kpr\right)$

with Kpr = private Capital and Kpub = Public Capital.

It shows that the different categories of public spending or public intervention are far from having the same effect on private sector productivity and overall growth. A 1% increase in public capital improves private sector productivity by 0.4%. He concludes that a deficit due to investment in economic and social infrastructure cannot be equal to a deficit due to the hiring of civil servants or environmental control. This study has the advantage of distinguishing in the empirical part the effect of capital expenditure and that of other public expenditure.

Artus and Kaabi (1993) carried out a study in OECD countries using a panel and found a positive and significant effect of public spending policy in “Research and Development” on the growth of GDP of those countries. Empirically, the use of a panel model is very suitable for this study, but there is the problem of stationarity and cointegration of panel data. In the Khan and Kumar (1997) model which covered a sample of 95 developing countries over the period from 1970 to 1990, it appeared that the effect exerted by public investments on production is lower than that exerted through private investments, which means that private sector investments are much more productive than public ones. The model used by Khan and Kumar (1997) is based on a Panel and makes it possible to assess the effectiveness of private and public sector investments. However, it does not integrate the role of time in economic analysis.

Amin (1998) carefully analyses the effects of fiscal policy on economic growth in Cameroon for the period from 1960 to 1994. Assuming that the GDP growth rate is obtained by aggregating investment growth rates public ( $Kpu{b}^{*}$), private investments ( $Kp{r}^{*}$), population ( $N^{*}$), accumulation of human capital ( $H^{*}$) and other factors of production ( $V^{*}$), Amin thus specifies the following model:

$Y_t^{*}=a_0+a_{1}Kp{r}_t^{*}+a_{2}Kpu{b}_t^{*}+a_3{N}_t^{*}+a_4{H}_t^{*}+a_5{V}_t^{*}$

The estimation of this model allows it to conclude that private and public investments have a positive and direct impact on economic growth. This model, however, raises the problem of cointegration between the different series of the model. In addition, all the variables in the model are dated at time t.

With the introduction of new growth models, empirical work has focused exclusively on the dynamic effect of fiscal policy on macroeconomic variables ( Kim, 1999 ). However, numerous empirical studies contribute to the “Fiscal Policy and Economic Growth” debate through time series models or panel models. In this section, we will distinguish works that present a positive relationship between fiscal policy and long-term growth from those that show a negative link.

Nairi et al. (2000) carried out empirical research in MENA countries by highlighting the relationships that exist between the components of fiscal policy variables and the level of GDP growth in these countries. They expressed the growth rate of GDP per capita (y) as a function of the following explanatory variables: the growth rate of lagged GDP per capita ( $y_{t-1}$), the ratio of private investment (RI) to GDP and the consumer price index (P), the ratio of current expenditure to GDP (RDC), the ratio of capital expenditure to GDP (RDEC), the ratio of budget balance to GDP (RSB). The basic MENA model is written as follows:

$\begin{array}{c}\Delta \ln y_t^i={\lambda }_0+{\lambda }_1\ln y_{t-1}^i+{\lambda }_2\ln RI{P}_t^i+{\lambda }_3\Delta \ln P_t^i+{\lambda }_4\ln RD{C}_t^i\\ +{\lambda }_5\ln RDE{C}_t^i+{\lambda }_6\ln RS{B}_t^i+{\lambda }_7\ln RR{B}_t^i+{\theta }_t^i\end{array}$

t and i represent the time and the country respectively; θ, being the error term.

These authors reached the following conclusions:

- Capital expenditure, budget revenues, and budget balance are positively correlated with the GDP per capita level. In addition, private investment has a positive influence on the level of growth;

- On the other hand, current public expenditure has a negative effect on the level of growth. This conclusion on current expenditure is also consistent with that of Barro and Sala-i-Martin (1992) .

The growth model presented by Nairi et al. (2000) on MENA countries reveals synchronous effects in the study because all the variables impact growth at the same period. In addition, this study introduces the variable “public consumption expenditure” which can only influence GDP growth in the short term.

Bose et al. (2007) carried out an analysis of panel data from 30 developing countries, and they explain the budgetary constraint as well as the possible biases due to the omission of variables in their model. At the end of their analyses, they show that there is a significantly positive relationship between the ratio of public capital expenditure and the growth of GDP per capita. In addition, he finds that the ratio of current budgetary expenditure has no effect on the level of growth of per capita income. A very striking result of this study is the insignificance of current public expenditure.

Through a study of 39 low-income countries, Gupta et al. (2005) showed that countries in which public expenditure is mainly made up of current expenditure (salaries, transfers) have very slow economic growth, while countries where capital expenditure accounts for a greater proportion of expenditure. Total state spending sees its economic growth faster.

Creel et al. (2015) show through a VAR model that public investment can have two effects on private investment: a crowding-out effect and a knock-on effect. They note that in France, public investment results in a ripple effect, while in the USA, the crowding-out effect is dominant, thus leading to a drop in private investment.

Sampognaro (2018) considers fiscal policy to be an effective tool to stop economic recessions and revive the economy. In the short term, he estimates that the value of the budgetary multiplier is close to unity and that in the long term, the effectiveness of budgetary policy depends on the economic context, in particular, the institutional context, the functioning of the goods market and factors of production, the degree of economic openness and the economic situation.

In his empirical work, Landau (1986) carried out a study on cross-sectional data from 104 countries and concluded that public consumption is negatively correlated with the level of growth per capita. Studies carried out by Barth and Bradley (1987) on 16 OECD countries between 1971 and 1983 also show a negative relationship between these same variables.

Work carried out by Barro (1989) on panel data from 98 countries reveals that public investment spending has no impact on growth; in addition, public administration consumption substantially reduces the level of GDP per capita. A fundamental limitation in this study is the insertion of the variable “Public consumption” in the estimation of the economic growth model. Indeed, consumer spending can only lead to a shift in the overall demand curve and not in the supply curve.

Easterly and Rebelo (1993) carried out analyses first on cross-sectional data on 100 countries between 1970-1988, then on panel data on 28 countries. At the end of their work, the following conclusions were drawn: investment spending on public infrastructure (public transport, communication) has a positive effect on GDP per capita, while capital spending, taken in its entirety, has a negative effect on per capita growth; the tax variables being determined by the model. Miller and Tsoukis (2001) extended the reasoning of Easterly and Rebelo and confirmed the hypotheses of these authors. The particularity of this study is that it shows that not all public capital expenditure is productive.

In the model of Dévarajan et al. (1996) , it appears from the analyses that the overall level of public expenditure has no impact on the level of growth, but by studying the effect of the components of public expenditure, there are some which have an impact on the level of growth. The model designed by these authors develops the link between the share of government spending and economic growth. The variable explained is real GDP per capita. Thus, at the end of their studies, they showed that current expenditure has a positive influence on economic growth while capital expenditure has a negative effect on the level of economic growth. This study poses a specification problem insofar as it assesses the impact of public consumption expenditure on long-term economic growth.

Temple (1999) , analysing the divergences in empirical conclusions, highlights a fundamental limitation of this work, namely the omission of the direction of state spending. He points out that the analysis of the relationship between fiscal policy and economic growth should not be reduced to the impact of public spending in its entirety; the increase in the latter may be a necessary and not sufficient condition. Thus, it can happen, and many developing countries tend to prove this, that a massive increase in state spending does not translate into productive investments and/or is financed in the form of projects with limited social returns, or even into “financial chasms” ( Schmidt-Hebbel et al., 1996 ).

According to Veroni and Saraceno (2005) , fiscal policy no longer explains the differences in terms of growth between the United States and the Eurozone. A budgetary shock of 1.2 points of GDP from 2001 to 2005 had a negative impact on the growth of the American economy. In the Eurozone, the abandonment of restrictive discretionary policy led to a deterioration of the budget deficit, while in the USA, it manifested itself as an interruption in the reduction of the public deficit.

Stiglitz (2016) shows in his work that certain Member States of the Eurozone initiated, after the 2007 financial crisis, a policy of budgetary austerity to revive their economies (Greece, Italy, Portugal) and avoid the spread of the crisis in the Eurozone. Their actions were supported by the Troika, as they were considered the best way to reduce the deficit through cuts in public spending and public debt. Four (4) years after the crisis in 2011, an assessment is necessary: the measures taken by the authorities certainly made it possible to avoid bank failure and the collapse of the financial markets. However, as for the rest, they were a real fiasco. Public debt has increased considerably and budget deficits have widened further, leading to an increase in social inequalities and low rates of economic growth.

For Bonfond (2016) , the observation was also the same in Belgium: cuts were made to social security spending and the administrations aimed to reduce the budget deficit and public debt to improve growth and, therefore, the competitiveness of the sector. The results were very disappointing following the budgetary austerity measures introduced by the Government after the crisis: growth remained very weak, poverty increased further and, worse still, public deficits and debt increased considerably.

Ngakosso (2018) , in addition to counter-cyclical and acyclic fiscal policies, the theoretical literature also identifies procyclical fiscal policy. Indeed, a fiscal policy is procyclical, if the public authorities increase (decrease) public expenditure and decrease (increase) the taxes in the expansion phase (recession). Such a fiscal policy strengthens the amplitude of the economic cycle, which is detrimental to inflation and employment.

Sampognaro (2018) notes a difference in the effects of fiscal policy in the short and long term. It contributes to the stabilization of the business cycle, but in the long term the values of the budgetary multiplier are weakened by the presence of non-Ricardian factors or monetary policy when the interest rate is at its minimum level (zero lower bound). In his work, he identifies the Eurozone, three fundamental phases of budgetary policy following the great economic recession of 2007 due to the subprime crisis:

- Phase 1: a very active budgetary policy to stop the recession and revive the economy from 2008 to 2010;

- Phase 2: budgetary consolidation covering the period from 2010 to 2015;

- Phase 3: neutrality of budgetary policy which results in primary structural balances close to zero.

International institutions such as the IMF, the OECD and the European Union agree on these three phases even if differences remain in the evaluation of potential GDP. In OECD countries, from 2008 to 2010, during the first phase, the primary structural balance stood at –3.4% of GDP. During the second phase, between 2011 and 2015, it became surplus with a very marked budgetary consolidation, bringing the primary balance to 4%. In the neutrality phase from 2016 to 2017, it is around −0.5%.

Kouika Bouanza and Ngassa (2021) state that the quality of governance in CEMAC needs to be improved to better contribute to the countercyclicality of public spending. This behavior contributes to economic stabilization and, thus, to sustained growth.

However, the economic literature presented in this section mentions differences in public spending. Indeed, public consumption spending is likely to only generate a shift in aggregate demand, leading to cyclical fluctuations, while public investments produce a shift in the aggregate supply curve, leading to long-term economic growth.

Even if the contribution of the public sector to economic growth is very low compared to that of the private sector, the objective of optimal growth must be the priority of the Government in the design and execution of budgetary policy. Since the objective is to achieve potential growth, the dependent variable of the model will be potential GDP ( $GD{P}^{*}$); which allows us to write the model as follows:

$GD{P}_t^{*}={\beta }_0+{\beta }_{1}dcon{s}_t+{\beta }_{2}de{c}_t+{\beta }_{3}dinv{p}_t+{\beta }_{4}E{C}_t+{\beta }_{5}TREN{D}_t+{\mu }_t$ (6)

with the long-term GDP at time t; $dcon{s}_t$, public consumption expenditure at time t; $de{c}_t$, public capital expenditure at time t; $dinv{p}_t$, private investment expenditures at time t and $E{C}_t={\left(dec\times dinvp\right)}_t$, the cross effect at time t between public capital expenditures and private investment expenditures.

The ${\beta }_i\left(i=1\text{to}5\right)$ represent the long-term multipliers.

Justification for the choice of variables

It is important to emphasize that, in the model, the dependent variable, long-term GDP (or potential GDP), is not a directly observable variable. Since the determination of potential GDP is not unanimous among economists, we will draw inspiration from the work of Nerlove (1981) who formulates the partial adjustment hypothesis according to which the variation of the dependent variable at time t is a fraction π of the desired variation for the same period. Which allows us to write:

$GD{P}_t-GD{P}_{t-1}=\pi \left(GD{P}_t^{*}-GD{P}_t\right)$ (7)

Two extreme cases may appear:

- If $\pi =1$, the current GDP equals the desired long-term GDP in the same period,

- If $\pi =0$, nothing changes since the effective GDP at time t is identical to that observed in the previous period. The coefficient π is expected to lie between these two extremes, because the adjustment of potential GDP to actual GDP is likely incomplete due to rigidities, inertia or contractual obligations; hence the name partial adjustment model. The adjustment mechanism of Equation (2) can be rewritten as follows:

$GD{P}_t=\pi GD{P}_t^{*}-\pi GDP+GD{P}_{t-1}$ (8)

$GD{P}_t=\pi GD{P}_t^{*}+\left(1-\pi \right)GD{P}_{t-1}$ (9)

Let’s replace it with his expression. We then have:

$\begin{array}{c}GD{P}_t=\pi [{\beta }_0+{\beta }_{1}dcon{s}_t+{\beta }_{2}de{c}_t+{\beta }_{3}dinv{p}_t+{\beta }_{4}E{C}_t\\ +{\beta }_{5}TREN{D}_t+{\mu }_t]+\left(1-\pi \right)GD{P}_{t-1}\end{array}$ (10)

$\begin{array}{c}GD{P}_t=\pi {\beta }_0+\pi {\beta }_{1}dcon{s}_t+\pi {\beta }_{2}de{c}_t+\pi {\beta }_{3}dinv{p}_t+\pi {\beta }_{4}E{C}_t\\ +\pi {\beta }_{5}TREN{D}_t+\pi {\mu }_t+\left(1-\pi \right)GD{P}_{t-1}\end{array}$ (11)

Equation (11) is called the short-term GDP function since, in this period, the effective GDP at time t is not necessarily equal to its long-term level. Once short-term Equation (11) has been estimated, we can easily derive the long-term GDP function and thus determine the long-term multipliers by dividing πβ_i by π.

Let us set:

${\Phi }_0=\pi {\beta }_0{\Phi }_1=\pi {\beta }_1{\Phi }_2=\pi {\beta }_2{\Phi }_3=\pi {\beta }_3{\Phi }_4=\pi {\beta }_4{\Phi }_5=\pi {\beta }_5\text{and}1-\pi =\alpha$

Model (6) becomes:

$\begin{array}{c}GD{P}_t={\Phi }_0+\alpha PI{B}_{t-1}+{\Phi }_{1}dcon{s}_t+{\Phi }_{2}de{c}_t+{\Phi }_{3}dinv{p}_t\\ +{\Phi }_4{\left(dec\times invp\right)}_t+{\Phi }_{5}TREN{D}_t+v_t\end{array}$ (12)

We are thus in the presence of a dynamic autoregressive model of order 1 which makes it possible to take into account the role of time in the analysis of economic decisions. The ${\Phi }_i\left(i=1\text{to}4\right)$ are the short-term multipliers which make it possible to measure the impact of the explanatory variables on the short-term effective GDP and α is the coefficient of the lagged endogenous variable. Once we have estimated the short-term multipliers, we will proceed to identify the long-term multipliers, which will make it possible to capture the effects of fiscal policy on long-term potential GDP.

Estimates are made on annual data ranging from 1961 to 2021, the data being available on the BCEAO website. Since the aim is to evaluate the sensitivity of GDP in relation to the different explanatory variables, the data are transformed into logarithms in order to have elasticities.

a) The model with overall public spending

The estimated model (in logarithm L) is written as follows:

$\begin{array}{c}LGD{P}_t=1.95+0.76LGD{P}_{t-1}+0.46LDP{T}_t+0.23LINV{P}_t\\ +0.017{\left(Ldec\times Linvp\right)}_t+0.09TREN{D}_t\end{array}$

The product of (Ldec × Linvp) corresponds to the cross effect (EC). The long-term elasticities are deduced from the short-term model estimates and the following Table 2 presents the short and long-term multipliers:

DPT = Log(Total public expenditure); INVP = Log(Private investment); EC = Cross-effect = DPT × INVP; TREND = Tendency. ^aAutocorrelation and normality tests of the residuals ensure that the residuals follow a white noise process. The models are generally significant and the risk threshold for the tests is 5%.

From these results, the following analyses emerge:

- Senegal’s current GDP depends 76% on the GDP of the previous year due to delays noted in the execution of economic policy decisions. Indeed, the implementation of economic policy decisions takes place with significant delays and the irreversible nature of certain measures taken by the government creates inertia in the management of public finances. Such behaviour impacts the time frame for adjusting actual GDP to potential GDP, which is 4 years and 2 months in Senegal. By applying the rule of 69, we see that to double Senegal’s economic growth rate, it would take 90 years, i.e. (69/0.76). This result is very understandable, especially since 1960, the economic growth rate of Senegal has been, on average, 3.22%. In 1961, the economic growth rate was 4.47% and 63 years later, Senegal is struggling to double its growth rate;

- The elasticity of effective GDP in relation to overall public expenditure in the short term is 0.46 and that of private investment is 0.23 in the short term. The two variables are significant in the model at 96% and 94% confidence levels respectively. We note that total public spending is very active within the framework of the economic recovery policy and its impact on effective GDP is higher than that of private investments which are more likely to lead to a shift in overall supply. In the long term, an increase in public spending of 1% leads to an increase in potential GDP of 1.91%. The implementation of the PSE for six (6) consecutive years from 2014 to 2020 resulted in a considerable increase in overall public spending and was distinguished by an average annual growth rate of 6.6%. After the COVID-19 shock, the Government put in place the Economic and Social Resilience Program with an envelope of 1000 billion CFA francs corresponding to a little more than 7% of GDP to allow the national economy to recover. Its potential growth trajectory;

- The estimation of the model shows us that the cross-effect is positive, but not significant. However, technical progress measured by the trend is significant in the short term, but weakly affects effective GDP with an impact of 0.0094. According to the United Nations (2016) , the low investment in research and development and the enrolment rate in higher education, combined with the inadequacy of economic policies and regulations unfavourable to progress, justify the low progress multiplier. Technology in least developed countries (LDCs).

To better understand the effects of fiscal policy, we will break down total public expenditure into public consumption expenditure and public capital expenditure.

b) The model with decomposed public expenditure

The estimated equation is given by:

$\begin{array}{c}LPI{B}_t=1.69+0.51LPI{B}_{t-1}+0.04LDCON{S}_t+0.62LDE{C}_t+0.33LINV{P}_t\\ -0.06{\left(Ldec\times Linvp\right)}_t+0.01TREN{D}_t\end{array}$

The cost and long-term multipliers are recorded in the following Table 3 :

DCONS = Log(public consumption expenditure); DEC = Log(public capital expenditure); INVP = Log(Private investment); EC = Cross-effect = DEC × INVP; TREND = Tendency.

It appears that the short-term budgetary multipliers are 0.04 for public consumption expenditure and 0.62 for public capital expenditure, respectively, with significance levels of 94% and 100%. These multipliers constitute the direct effects of fiscal policy on short-term economic growth. These results demonstrate the effort of public authorities in terms of public investment. By wanting to place Senegal on the trajectory of emergence by 2035, the authorities have strengthened public investments in agriculture, economic and social infrastructure (roads, airports, road tolls, hospitals, universities, energy, etc.).

The impact of private investment on short-term growth improved in the last model (+0.33) and still retains its positive sign. The particularity of this model is the significance of the crossed effect with a negative multiplier of −0.06. This allows us to highlight two effects of budgetary policy focused on public investments: a direct effect and an indirect effect:

* in the short term, public investments have a direct impact on effective growth of 0.62% and the indirect effect is negative at 0.06%. Thus, the total effect of public investments on effective GDP in the short term is +0.56%.

* in the long term, the total impact of public investments on potential GDP is 1.14% with a long-term crowding out effect of −0.14%. Public investments in Senegal result in crowding out the private sector, particularly due to the financing of its investments with internal resources. This crowding-out effect thus reduces the effectiveness of Senegal’s budgetary policy in the short and long term, weakening the value of the total multiplier. The first phase of the PAP (Priority Action Plan) benefited from funding from external resources for 38%, from internal resources for 54% and from the private sector for 7%. These efforts by the Government to finance the PAP2A have limited the participation of the private sector in the dynamics of economic growth.

What lessons can be learned from these results in relation to the adjustment of public finances which is expected in this context marked by non-compliance with budgetary rules?

Estimates show that the Government is running enormous risks by making budget cuts on its investments which are easier to reduce to restore balance to public finances. A reduction in public investments of 10% in Senegal translates in the short term into a drop in effective GDP of 6.2% and a reduction in long-term potential GDP of 11.4%. Through these results, it becomes easy to understand that the objective of economic efficiency and growth can itself lead to problems of arbitration between investment and consumption. As part of the economic recovery policy, the authorities are placing more emphasis on public consumption spending to quickly create a “Go” effect, because public investments are executed with many delays.

The analysis of the effects of budgetary policy on economic growth should not be reduced to the impact of public spending in its entirety. To achieve the objectives set within the framework of the PSE while taking into account the vulnerable shocks likely to affect the national economy, it is essential for the authorities to give an important place to the orientation of public expenditure, especially in a context marked by non-compliance with budgetary rules (public debt greater than 70% of GDP and a deficit greater than 3% of GDP). In a phase of budgetary consolidation, it is essential to carefully target the budgetary expenditures to be reduced so as not to compromise actual and potential economic growth. The results of our estimates show us above all that public capital expenditure has a greater impact on economic growth, despite its execution time. However, in a period of consolidation of public finances, Governments are more tempted by a reduction in public capital expenditure than by a reduction in consumer expenditure to ensure the balance of public finances. State consumption expenditure is irreversible and its reduction can raise tensions with unions or social organizations, unlike public investments which are not defended by pressure groups. Governments can suspend them at any time.

As part of the diagnosis of national accounts as carried out by the IMF, it must be emphasized that the positive relationship between public spending and economic growth through the analysis of contributions to growth could thus come from the identity effect accounting which means that, when a component of overall demand (such as public investments) increases, GDP changes in the same direction. Such results must be interpreted with great caution, because such an analysis does not make it possible to highlight the direct and indirect effects of budgetary policy on the level of economic activity and on long-term potential growth.