The Impact of Tax Pressure on Long-Term Economic Growth in Morocco

Abstract

Despite the tax reform in the 1980s, the Moroccan tax system still suffers from structural vulnerabilities that have led to inefficient tax policies and increasing sectoral disparities, which have led to an increase in the tax pressure rates between 1990 and 2020. To overcome these vulnerabilities, and according to the recommendations of the 2019 National Tax Conference, the Moroccan tax system has been recently restructured, with the aim of strengthening the principles of equity and tax efficiency. The main objective of this paper is to determine the impact of the tax pressure on long-term economic growth in Morocco from 1990 to 2020. To do so, the methodology adopted in this work consists of verifying the effect of the tax pressure on economic growth through a quantitative methodology based on two vector autoregression approaches: the vector autoregression model (VAR) and the vector error correction model (VECM). The results of this study confirm that the effect of fiscal pressure on economic growth in Morocco is not significant in the long run, and therefore, we can deduce that taxation is not yet a well-mastered instrument for the state to act on the economic sphere. The results show a complex relationship between tax pressures and economic growth in Morocco, underlining the importance of tax reform. However, the research is limited by the specific models and the need to explore other determinants. The contribution of this paper lies in its in-depth empirical analysis of fiscal pressure and its influence on long-term economic growth in Morocco, as based on VAR and VECM modeling. This econometric approach makes it possible to isolate the dynamic effects and responses of economic variables over time, offering an understanding of the interactions between tax policy and long-term economic growth. By focusing specifically on the Moroccan context, this study offers an original perspective and helps to fill the gap in the empirical research in this area while providing a valuable analytical framework for assessing fiscal policies and their impact on long-term economic growth.

1. Introduction

Zagler and Dürnecker (2003) state that tax policy plays a key role in shaping a country’s economic growth. Decisions about government spending and revenues have a direct impact on the pace of growth. Expansionary fiscal policies, such as tax cuts or increases in public spending, can stimulate aggregate demand and boosting growth, especially during a recession. Conversely, restrictive fiscal policies can control inflation and stabilize the economy in periods of overheating. A well-designed tax policy also helps to reduce social and territorial inequalities, creating an environment conducive to more inclusive and sustainable economic growth. Understanding the influence of tax policy on economic growth provides decision-makers with the tools they need to design effective economic policies tailored to the specific needs of their country.

In this regard, an examination of the objectives of tax policy prompts us to question the degree of influence exerted by the tax parameter on the achievement of economic growth. Consequently, this places us in a position to assess the level of tax pressure required by the state to achieve the objectives of optimal long-term economic growth.

In response to Morocco’s public finance crisis during the 1980s, the government undertook a fiscal policy reform under the Structural Adjustment Plan (SAP). The aim was to establish a modern, coherent and efficient fiscal and economic policy that would provide sufficient financial resources to overcome the challenges posed by the crisis. The main focus of this reform, through Framework Law No. 3.83 on tax reform, was to replace cedulas with synthetic taxes, thereby increasing the number of taxpayers and tax revenues, broadening the tax base, and minimizing the tax pressure.

Despite the implementation of this tax reform, the Moroccan tax system still suffers from structural vulnerabilities that have led to inefficient tax policies and increased sectoral disparities, resulting in higher rates of tax pressure between 1990 and 2020. To overcome these vulnerabilities, in line with the recommendations of the 2019 National Tax Conference, the Moroccan tax system was restructured through Framework Law No. 69.19 on Tax Reform Fiscal (2021), aiming to reinforce the principles of fairness and tax efficiency.

The homogeneous, stable and chronically uneven distribution of tax pressure levels in Morocco confirms that the Moroccan tax system is heavily reliant on consumers and wage earners, leading to a reduction in labor income in favor of capital, squeezing consumption and domestic demand.

Based on studies already carried out in Morocco, such as the work of Afifi and Ramdaoui (2019) on the relationship between tax pressure in Morocco and economic growth by estimating the Scully model, Salhi and Echaoui (2020) have developed a study to model the optimal tax rate in Morocco over the period 1985–2019, and Belahouaoui and Attak (2022) analyzed the relationship between tax pressure and taxpayer compliance through a review of the theoretical literature. It becomes evident that there is a dearth of studies on tax pressure and its impact on the long-run economic growth in Morocco. In this sense, the contribution of this article lies in its in-depth empirical analysis of fiscal pressure and its influence on long-term economic growth in Morocco, as based on VAR and VECM modeling. This econometric approach makes it possible to isolate the dynamic effects and responses of economic variables over time, offering an understanding of the interactions between tax policy and long-term economic growth. By focusing specifically on the Moroccan context, this study offers an original perspective and helps to fill the gap in the empirical research in this area while providing a valuable analytical framework for assessing fiscal policies and their impact on long-term economic growth.

The aim of this article is to determine the impact of tax pressure on long-term economic growth in Morocco. The problem posed in this article is to answer the following question: what is the impact of tax pressure on long-term economic growth in Morocco?

The methodology adopted in this work consists of verifying the effect of tax pressure on economic growth using a quantitative methodology based on two vector autoregression approaches: the vector autoregression model (VAR) and the vector error correction model (VECM).

This article will be subdivided into four parts: the first deals with a review of the theoretical literature; the second is devoted to the presentation of stylized facts related to the evolution of tax pressure rates in Morocco; the methodology of the work is presented in the third part; and the last part is devoted to the results and discussion.

2. Literature Review

2.1. Taxation and Economic Growth

It is important to note that taxation cannot be seen as a universal solution to all economic problems and that it needs to be designed and used strategically to maximize its impact on growth. Furthermore, it is essential to consider the long-term economic and social consequences of taxation, as a tax that may appear beneficial in the short term may have negative effects in the long term.

Keynes (1936) advocated the role of government intervention through fiscal policy in influencing economic activity. He also supported the ideology that aggregate demand stimulates an economy’s level of production. According to Keynesian theory, taxation influences the level of disposable income and aggregate demand, thereby affecting the quantity of production.

Solow (1956) has called into question the Harrod–Domar model of economic growth. This model concludes that the economy is in an unstable long-term growth equilibrium based on key parameters such as the savings rate, the capital–output ratio and the rate of increase in the working population. However, Solow showed that this conclusion was largely based on the fundamental assumption that production takes place under conditions of fixed proportions. By abandoning this assumption, the concept of unstable equilibrium seems to lose its relevance. It is therefore essential to examine tax structures carefully so as to minimize the negative impact of taxes on economic growth while preserving budget revenues.

The exogenous growth model underlines the risks of excessive taxation, noting its negative impacts on investment, innovation and economic activity, while accentuating the complexity of the relationship between taxation and external growth.

In short, taxation can have a significant impact on exogenous economic growth, but it is important to consider the complexity of this relationship and to design taxation adapted to the economic and social circumstances of each country.

In the 1980s, with the advent of neo-liberalism, the classical vision of taxation returned with a vengeance, with reduced taxation and increased privatization of public services. This approach was criticized for its impact on social inequality and its lack of support for long-term growth.

Romer (1990) explored the concept of endogenous economic growth. This model emphases the essential role of human capital and intentional investment in research and development in stimulating economic growth. Thus, the relationship between economic growth and taxation is complex, as tax policies can influence these investment decisions and, consequently, long-term growth.

The endogenous model accentuates the impact of corruption on the ideal taxation level by predicting an inverted U-shaped relationship between taxes and growth. These viewpoints come together to emphasize the significance of carefully considered taxation that can support wise public investment. They also emphasize how equitable taxation has the power to lessen inequality, enhancing social cohesiveness and promoting sustained economic expansion. To put it briefly, these models highlight the necessity of a just tax system in order to support long-term economic growth.

The model of Barro et al. (1990) on growth and taxation is an economic theory that suggests that high levels of taxation can reduce long-term economic growth by discouraging investment and capital accumulation. According to this model, governments should seek to maintain a balance between public spending and tax revenues to maximize long-term economic growth. However, this model is criticized for failing to take into account factors such as externalities and inequalities, which can have a significant impact on the effectiveness of tax policy.

Endogenous growth economists have shown that taxation can influence economic growth by affecting human capital accumulation, innovation, research, development and infrastructure investment. Tax policies that favor these factors can stimulate economic growth in the long term, while inadequate tax policies can slow it down. However, the extent of the impact of taxation on economic growth depends on many other economic and institutional factors, and economists do not all agree on the nature and importance of these effects.

In recent decades, the relationship between taxation and economic growth has been studied in greater depth, and it has become clear that the issue is complex and that there are no simple answers. Economists have stressed the importance of designing and implementing tax policies that support long-term economic growth while preserving equity and macroeconomic stability.

Engen and Skinner (1996) have researched the link between taxation and economic growth. Their main conclusion is that taxation can have a significant impact on economic growth but that this impact depends on the type of tax and how tax revenues are used. They found that taxes on labor and capital can reduce investment and employment, which can have a negative effect on economic growth. Conversely, taxes on consumption and property have less of a negative effect on growth. Engen and Skinner also found that the way tax revenues are used can have a significant impact on economic growth. For example, if tax revenues are used to finance inefficient public-spending programs, this can slow economic growth. On the other hand, if tax revenues are used to finance investment in education, infrastructure or research, this can stimulate economic growth.

In summary, according to Engen and Skinner (1996), a well-designed tax policy can promote economic growth by avoiding taxes on labor and capital, making judicious use of tax revenues and reducing inefficient public spending.

In his empirical analysis of the relationship between tax pressure and economic growth, Scully (2000) has conducted studies suggesting that high levels of tax pressure have a negative effect on economic growth. His research examines the relationship between taxation, economic growth and income inequality. He explores the optimal tax policy that can promote economic growth while reducing income inequality. The author argues that the tax system can have both positive and negative effects on economic growth and income inequality. While taxation can be used to finance public programs and services, it can also discourage work and investment, which can hinder economic growth.

According to Scully (2003), optimal tax policy must balance the need for revenue with the need to promote economic growth and reduce income inequality. He suggests that a progressive tax system, which taxes higher incomes at a higher rate, can help reduce income inequality without significantly harming economic growth.

However, the research also indicates that tax policies need to be carefully designed to avoid unintended consequences, such as reducing incentives to work or creating tax loopholes for the wealthy.

Overall, Scully’s model highlights the complex relationship between taxation, economic growth and income inequality, and it suggests that optimal tax policy requires careful consideration of these factors.

Lee and Gordon (2005) show that corporate tax rates are significantly negatively correlated with differences in economic growth rates between countries, even after controlling for other determinants of growth. In addition, fixed-effects regressions confirm that increases in corporate tax rates lead to lower future growth rates within countries.

Arnold (2008) suggests that the composition of taxes has a significant effect on economic growth. More specifically, property taxes, particularly recurring taxes on real estate, appear to be the most growth-friendly, followed by consumption taxes and then personal income taxes. By contrast, taxes on corporate profits have the most negative effect on the GDP per capita. These findings suggest that a growth-oriented tax reform should favor property and consumption taxes over income taxes, particularly corporate taxes.

The relationship between taxation and economic growth is a complex and multidimensional area. Tax policies, such as tax rates, the structure of taxes and their impact on incentives to work, invest and consume play an important role. Studies show that property and consumption taxes promote economic growth, while taxes on corporate profits have a negative effect on the GDP per capita.

Edame and Okoi (2014) show that taxation has an inverse relationship with investment: a one percentage point increase in corporate income tax leads to a decrease in the level of investment in Nigeria. Furthermore, taxation is positively related to public spending. These findings underline the importance of balanced tax reform in fostering economic growth and development.

Takumah and Njindan Iyke (2015) posit that the theoretical underpinnings of taxation and economic growth suggest negative and positive correlations between these two variables. The negative correlation is due to the distortions and suppression inherent in taxes, while the positive correlation is indirectly due to tax-financed spending.

Alinaghi and Reed (2021) state that the relationship between taxation and economic growth is complex and varied. While high taxes can reduce incentives to work, invest and consume, they can also fund public goods and services that encourage economic development and citizen well-being. To promote sustainable growth, it is essential to design the tax system astutely so as to minimize the negative impact of taxation on growth while preserving tax revenues.

To conclude this section, taxation, as a tool for financing public spending, plays an essential role in economic development. Theoretical models emphasize the potential impact of taxes on investment, innovation and long-term growth. However, the empirical reality is more complex, as it depends on the composition of taxes, their use and the specific context of each country. Effective tax reform, supported by sound public policies, is essential to promote long-term economic growth. Moreover, the relationship between tax pressure and economic growth is complex and can vary according to economic conditions, the structure of the economy and the quality of public spending. It is therefore important to consider the potential effects on growth when designing tax policies.

2.2. The Theoretical and Empirical Relationship between Economic Growth and Tax Pressure

The relationship between tax pressure and economic growth is a controversial topic in economics. On the one hand, high tax pressure can dampen economic growth by reducing the availability of capital for productive investment and discouraging companies and individuals from earning and investing. On the other hand, lower tax pressure can stimulate growth by encouraging investment and consumption.

The Laffer (1981) curve was modeled in 1974 by supply-side theorist Arthur Laffer and reflects the liberal conception of redistribution as inefficient since it reduces the incentive to work and invest. This disincentive can be explained by the role of compulsory deductions, the Laffer curve also shows the negative effect of a high rate of compulsory deductions. It is an old idea, as Smith (1776) states, “taxation can hinder industry in the sense of labor”.

For this theory, Adam Smith and Jean Baptiste Say have already spoken on the basic premise of Ibn Khaldun’s writings, based on his idea that “too much tax kills tax. Adam Smith and Jean Baptiste Say said that “an exaggerated tax destroys the base on which it is levied, and that a reduction of the tax increases the revenue of the State, so that governments can be sure that it is better to be moderate”.

Laffer then highlights a relationship between two variables, the first being tax revenues, and the second being the tax rate, i.e., the share of taxes or compulsory tax deductions in national wealth; in other words, the share of wealth deducted by the state in the broad sense.

The curve shown in Figure 1 illustrates the relationship between the tax rate and tax revenue. On the vertical axis (ordinate) is the tax revenue, while the horizontal axis (abscissa) represents the tax rate. Initially, a low tax rate, represented by point A on the curve, is associated with limited tax revenue. As the tax rate increases, tax revenue continues to rise until it reaches the optimal tax rate, marked by point E on the curve. Beyond point E, any further increase in the tax rate results in a decrease in tax revenue. Therefore, the same amount of tax revenue can be achieved at different levels of the tax rate. This relationship is explained by the following figure:

It seems logical that if the tax rate is limited, the amount of tax revenue will be limited, so increasing the tax rate results in an increase in tax revenue, but the closer you get to the optimum tax rate, the more limited the increase in tax revenue. On the other hand, beyond this optimum rate, increasing the tax rate results in a drop in the tax revenue, with prohibitive rates and perverse effects.

Firstly, households have less incentive to work and make an effort, so they are convinced that part of the wealth they have created is being taken away by the state. This can lead some households to immigrate to countries with lower taxes (tax havens), which is known as tax evasion. Secondly, entrepreneurs will move abroad in search of more lenient taxation (tax exits). The combination of these two factors will lead to a reduction in wealth and, therefore, have a negative impact on economic growth, if not cause a reduction in economic growth.

The conclusion that can be drawn from this theoretical part is that tax pressure has a considerable impact on economic growth, but what is the impact of economic growth on tax pressure? A number of empirical studies of the relationship between fiscal pressure and long-term economic growth have subsequently been carried out. The results obtained differ from country to country, depending on the variables used and the methodology employed to model this relationship.

Van Heerden and Schoeman (2008) assessed the optimal size of government in terms of revenue and expenditure for South Africa in order to maximize economic growth, using time-series data for the period 1960 to 2006. Van Heerden and Schoeman (2008) showed in their study that South Africa’s average tax pressure may be on the downward side of the Laffer curve. Consequently, the tax pressure has a negative impact on economic growth.

Keho (2010) estimated the optimal tax pressure for the Ivorian economy using both the Scully and quadratic regression models using time-series data between 1960 and 2006. The results are not consistent with the proposition that higher tax rates are detrimental to growth.

Takumah and Njindan Iyke (2015) explored the causal influence of tax revenues on economic growth in Ghana over the period 1986–2014. The results of this research provided strong evidence of a unidirectional causal flow of tax revenues to economic growth in Ghana. These results are in line with existing findings that taxation can influence economic growth. The policy implication is quite clear.

Saibu (2015) estimated the optimal tax rates for South Africa using quarterly data for the period 1994–2016, employing an ARDL bounds testing approach. The results revealed that there is no significant relationship between taxation and economic growth over the period studied.

Chokri et al. (2018) have attempted to study the relationship between the level of tax pressure and the growth rate for the Tunisian case with the estimation of the optimal tax rate between 1966 and 2015. The methodology adopted in this study consists of two stages. In the first stage, the optimal tax rate is determined using Scully’s static model and the quadratic model. The second stage focuses on the long-term relationship. This research is based on the results of the unit root and cointegration tests on the Scully model. The results of the study support the idea that taxes reduce growth beyond a certain threshold. The basic model yielded an optimal tax rate equal to 19.6% of the GDP. The use of time-series cointegration techniques enabled a long-term analysis to be carried out, yielding an optimal tax rate of 14% of the GDP. The current tax rates are well above these levels, which explains the disappointing performance in terms of growth and taxation.

Koatsa et al. (2021), in their work, tried to determine whether tax pressure has an impact on the economic growth rate in the Lesotho context and to estimate an optimal tax pressure over the period 1988–2017. The ARDL bounds test framework was used to establish cointegration and determine whether there is a long-term relationship between tax pressure and economic growth. The results established a cointegration relationship between economic growth rate and fiscal pressure, with unidirectional causality running from economic growth to fiscal pressure. Granger causality revealed no causal effect between fiscal pressure and economic growth in Lesotho. The results show that tax policy has had no significant influence on economic activity in Lesotho over the period studied. However, the results of the error correction model indicate a long-term relationship between the economic growth rate and tax pressure in Lesotho, with a long-term equilibrium adjustment speed of 100% following a short-term shock.

Based on these studies, we can see that there is still no consensus on the relationship between tax pressure and economic growth.

An initial series of studies by Van Heerden and Schoeman (2008) in South Africa and Keho (2010) in Côte d’Ivoire found that tax pressure has a negative impact on economic growth, while a second series of studies by Takumah and Njindan Iyke (2015) in Ghana found that taxation can influence economic growth, and a third study by Saibu (2015) in South Africa confirmed the findings of Van Heerden and Schoeman (2008) and showed that there is no significant relationship between tax pressure and economic growth. Chokri et al. (2018) found that in Tunisia, an increase in tax rates explains a disappointing performance in terms of taxation and economic growth. The latest series of studies by Koatsa et al. (2021) in Lesotho showed that in the short-term, tax pressure does not influence economic growth, but on the basis of the error correction model, there is a long-term relationship between tax pressure and economic growth.

This last study is like the objective of this empirical work, which is to study the impact of fiscal pressure on long-term economic growth in Morocco.

Assumptions:

H0. 

Tax pressure has an impact on long-term economic growth.

2.3. Tax Pressure and Economic Growth in Morocco

The Moroccan government’s tax revenues come from a number of taxes. However, the structure of this revenue shows that Value Added Tax (VAT), Income Tax (IC), Corporation Tax (CT), Registrations and Stamps and Internal Consumption Tax (ICT) are the main sources of tax revenue for the state (Figure 2).

According to Figure 2, VAT, CT, IT and registration fees are the main tax revenues in the structure of state tax revenues, with an average of 26.53% for VAT revenues, 22.1% for CT, 19.9% for IT and 14.77% for registration and stamps between 2000 and 2020.

During the period 1990–2020, the tax pressure and the growth rate of the real GDP as an annual percentage in Morocco have undergone profound changes according to data collected from the International Monetary Fund, as shown in the following graph:

During the period 1990–2000, Morocco undertook a series of political, social, financial and economic reforms, as well as implementing the Framework Law on taxation and reforming the national tax system through the Finance Acts during this period, with the aim of catching up on accumulated delays in economic and social development. According to Figure 3 and Figure 4, tax pressure in Morocco underwent various changes over the period 1990–2000, decreasing by 0.48% from 21.26% in 1990 to 20.78% in 2000, i.e., an average tax pressure of 20.15% between 1990 and 2000, with an economic growth rate of 3.15% over the same period.

The first “Taxation Assizes”, held in 1999, provided an opportunity to reflect on the mechanisms for simplifying, streamlining and harmonizing the tax system. Numerous reforms were introduced by successive Finance Acts from 2000 to 2011, resulting in a series of measures to simplify, rationalize and harmonize the tax system, including the reform of registration duties in 2004, the start of VAT reform in 2005, the consolidation of tax texts into a single volume, the General Tax Code published in 2007, and the reform of the income tax scale in 2009, adding to the financial crisis that hit the world in 2007 and 2008. Over the period 2000–2010, the rate of tax pressure rose sharply from 20.78% in 2000 to 24.78% in 2010, an increase of 4 points, with the highest rate recorded in 2008 at 28.88%, while the geometric average rate of tax pressure over the period 2000–2010 was 23.24%, an increase of 3.09% on the 1990–2000 period, with an average economic growth rate of 4.68%. Despite the increase in tax pressure between the two periods, the economic growth rates themselves increased by 1.53%, a finding that contradicts the data in the theoretical literature review.

Since 2010, Morocco has launched a series of structural reforms (the 2011 constitution, the organic laws governing local authorities, the organic finance law, the tax framework law following the recommendations of the national tax conference in 2019, and the reforms introduced in the wake of the COVID-19 crisis). The rate of tax pressure increased from 24.78% in 2010 to 27% in 2020, giving a geometric average rate of 24.99% of tax pressure over the 2010–2020 period, while the average rate of economic growth fell from 4.68% over the 2000–2010 period to 2.54% over the 2010–2020 period.

Since 2020, due to the COVID-19 pandemic, which triggered an economic crisis, global economic growth has plummeted from 3% to −6.3%. To mitigate the adverse effects of this crisis, governments and central banks have implemented fiscal and monetary measures on an unprecedented scale (Benmelech and Tzur-Ilan 2020).

In Morocco, the crisis significantly impacted key sectors of the economy (tourism, transportation, textiles, and industry, among others), causing the growth rate to decline from an average of 4.1% between 1999 and 2019 to −7.2% in 2020. This critical situation prompted the Moroccan government to intervene through budgetary, monetary, and fiscal policies aimed primarily at mitigating the effects of the economic crisis. The measures included the establishment of a Special Fund for managing the COVID-19 pandemic to provide direct assistance and subsidies to vulnerable households and businesses, the postponement of financial and tax deadlines, the adoption of a supplementary budget in 2020, and the reduction of the policy interest rate to as low as 1.5%, supplemented by the activation of a USD 3 billion liquidity line from the International Monetary Fund (IMF).

Moreover, the COVID-19 crisis accelerated reforms to the Moroccan tax system between 2020 and 2023. The key measures included the adoption of Framework Law No. 69.19 on Tax Reform Fiscal (2021), the creation of the Unique Professional Contribution (CPU), reductions in corporate income tax (CIT) and Value Added Tax (VAT), as well as exemptions and tax deductions for sectors affected by the COVID-19 pandemic.

Although the Moroccan government’s intervention significantly mitigated the economic effects of the COVID-19 health crisis, it also led the national economy, similar to developed and developing economies, into an inflationary spiral (from 0.7% in 2020 to 6.7% in 2022), high fiscal pressure (27.3% in 2020 to 22.6% in 2023), a substantial budget deficit (7.5% of GDP in 2020), and a significant decline in tax revenues of more than 79%. This situation aligns with the findings of Silva (2021), suggesting that expansionary fiscal policies increase government deficits, which in turn affect the profitability of the banking sector. Furthermore, high inflation exacerbates fiscal deficits, given the risks of fiscal dominance.

On a panel concerning Morocco and 33 OECD African countries, the OCDE (2023) stated the following:

Between 2010 and 2021, the increase in the average tax-to-GDP ratio in Africa, as presented in Figure 5, was mainly attributable to higher VAT and personal income tax revenues. In 2021, taxes on goods and services remained the main source of tax revenue in Africa, accounting for 51.9% of the total tax revenue on average, with VAT contributing 27.8%. Taxes on income and profits provided 37.9% of tax revenues. For 24 of the African countries included in this report, taxes on goods and services were the primary source of tax revenue, while in nine other countries, taxes on income and profits predominated.

Between 2020 and 2021, revenue from taxes on goods and services rose by an average of 0.2% of the GDP, representing a modest rebound after a fall of 0.4% of the GDP in 2020, attributable to the impact of the COVID-19 pandemic. In this category, VAT receipts rose by 0.1 percentage points in 2021, marking a slight rebound after a 0.3 percentage point decline between 2019 and 2020. Income tax receipts fell slightly by 0.1 points on average in 2021, mainly due to lower corporation tax receipts over this period, after remaining stable between 2019 and 2020. Social security contributions fall by 0.1 points in 2021, following a similar increase in 2020.

3. Data and Methodology

3.1. Data

This study uses annual time series for a 30-year period from 1990 to 2020. The choice of this period is mainly motivated by data availability and the structural changes that Morocco has undergone over the years (The data used are limited to 1990–2020, because the tax reform in Morocco took place in 1983, giving rise to the main taxes, so data availability only concerns the period studied). The data sample is based on statistics drawn from the International Monetary Fund (IMF) database for the growth rate of the real GDP and Research and Financial Forecasting Department of Moroccan Ministry of the Economy and Finance for the pressure tax rate. The aim is to examine the causal relationship and the impact of economic growth on fiscal pressure in Morocco.

It is important to emphasize that the period of the COVID-19 crisis was not included in this analysis due to its exceptional and unusual characteristics. Given that our study focuses on long-term trends, this exclusion was necessary to ensure the relevance and consistency of the results.

To this end, the econometric models in this article deal with two variables of interest. On the one hand, the growth rate of the real GDP as a dependent variable that measures economic growth relative to the gross domestic product (GDP) from one period to the next, adjusted for inflation and expressed in real terms as opposed to nominal terms. On the other hand, the tax pressure rate as an independent variable remains the most widely used indicator for determining the weight of taxation in an economy. Similarly, this rate reflects tax policy and provides information on the totality of the tax measures undertaken by the state, while taking into account the collection capacity of the tax administration.

It should be noted that the calculation of tax pressure here only covers tax revenues and not all compulsory levies, notably social security contributions, as these are considered to be resources earmarked for redistribution operations, and their purpose is to modify the distribution of income. On the other hand, in this study, we will examine taxes as such, i.e., all the direct and indirect contributions collected by public administrations, excluding social contributions.

The choice of variables adopted to study the relationship between tax pressure and economic growth in Morocco draws on the work of Van Heerden and Schoeman (2008), Takumah and Njindan Iyke (2015), Saibu (2015), Chokri et al. (2018), and Koatsa et al. (2021), limited to the growth rate of the real GDP and the rate of tax pressure excluding social contributions. This study focuses solely on the impact of direct and indirect taxes on economic growth.

For a comprehensive understanding of the variables,

Table 1. Definitions and data sources.

Variable	Abbreviation	Measure	Source
Tax pressure rate	PF	${\displaystyle \frac{Totaltaxrevenues}{GPD}}\times 100$	Research and Financial Forecasting Department (2008), Elbaggari (2023)
Growth rate of real GDP	GPD_t	${\displaystyle \frac{{GPD}_t-{GPD}_{t-1}}{{GPD}_{t-1}}}\times 100$	International Monetary Fund (IMF): https://www.imf.org/en/Countries/MAR, accessed on 28 March 2022

Source: Authors.

below provides definitions and data sources.

3.2. Methodology

3.2.1. Model Selection

The theoretical aspect of this research was complemented by an empirical study. This was carried out using a quantitative methodology based on two approaches to vector autoregression: the vector autoregression model (VAR) and the vector error correction model (VECM).

Vector autoregressive modeling (VAR) aims to describe the interdependencies between all the variables. In fact, the generalization of the VAR representation to k variables and p shifts (noted VAR(p)) is written in matrix form:

$${\mathit{Y}\_}_{\mathit{t}}={\mathit{A}\_}_{\mathbf{0}}+{\mathit{A}}_{\_\mathbf{1}}{\mathit{Y}\_}_{\mathit{t}-\mathbf{1}}+\dots +{\mathit{A}\_}_{\mathit{p}}{\mathit{Y}\_}_{\mathit{t}-\mathit{p}}+{\mathit{V}\_}_{\mathit{t}}$$

(1)

$$\mathrm{With}{\mathit{Y}\_}_{\mathit{t}}=\left(\begin{array}{c}{\mathit{Y}}_{\mathbf{1},\mathit{t}}\\ ..\\ {\mathit{Y}}_{\mathit{k},\mathit{t}}\end{array}\right);{\mathit{A}\_}_{\mathit{i}\ne \mathbf{0}}=\left(\begin{array}{ccc}{a}_{1i}^1& a_{1i}^2...& a_{1i}^k\\ a_{2i}^1& a_{2i}^2...& a_{2i}^k\\ a_{ki}^1& a_{ki}^2...& a_{ki}^k\end{array}\right);{\mathit{A}\_}_{\mathbf{0}}=\left(\begin{array}{c}{a}_1^0\\ a_2^0\\ \begin{array}{c}..\\ a_k^0\end{array}\end{array}\right);{\mathit{V}\_}_{\mathit{t}}=\left(\begin{array}{c}{\mathit{V}}_{\mathbf{1}\mathit{t}}\\ \begin{array}{c}{\mathit{V}}_{\mathbf{2}\mathit{t}}\\ ..\\ {\mathit{V}}_{\mathit{k}\mathit{t}}\end{array}\end{array}\right)$$

Of which:

Y_t = (${\mathit{Y}}_{\mathbf{1},\mathit{t}}$, ${\mathit{Y}}_{\mathbf{2},\mathit{t}}$, ..., ${\mathit{Y}}_{\mathit{k},\mathit{t}}$), a vector of k endogenous variables introduced into the system, where each variable constitutes an equation.

${\mathit{A}\_}_{\mathit{i}\ne \mathbf{0}}$ is the matrix of coefficients of order k × k to be estimated.

${\mathit{V}\_}_{\mathit{t}}$ = (${\mathit{V}}_{\mathbf{1}\mathit{t}}$, ${\mathit{V}}_{\mathbf{2}\mathit{t}}$, ..., ${\mathit{V}}_{\mathit{k}\mathit{t}}$), vector of shocks assumed to be non-autocorrelated white noise of constant variance.

The aim is to understand the dynamic behavior of variables that are linearly dependent on the past. Moreover, by considering the relationships between several variables, the model makes it possible to explain, not just describe, the evolution of a series. The method for estimating the VAR model is as follows. First, study the stationarity of the variables, then determine the number of lags (p) using the information criteria to determine the optimal VAR. The latter must be validated by the significance of the coefficients and the analysis of the residuals.

It is also necessary to specify the model in econometric form. Ultimately, this model will take the following simple form:

$${\mathbf{G}\mathbf{D}\mathbf{P}\_}_{\mathit{t}}={\mathsf{\alpha }\_}_{\mathbf{0}}+{\mathsf{\beta }}_{\_\mathbf{1}}{\mathit{P}\mathit{F}}_{\_\mathit{t}-\mathbf{1}}+{\mathsf{\epsilon }\_}_{\mathit{t}}$$

(2)

Of which:

• ${\mathbf{G}\mathbf{D}\mathbf{P}\_}_{\mathit{t}}$ The real economic growth rate in period t
• $\mathbf{P}\mathbf{F}$ Tax rate.
• ${\mathsf{\alpha }}_{\mathbf{0}}$ Model constant.
• ${\mathsf{\epsilon }\_}_{\mathit{t}}$: Model error: variables not considered in period t.

In fact, the VAR model alone cannot measure the relationship between economic growth and tax pressure, which is why it is often combined with the vector error correction model (VECM). This model examines short-term causality as well as long-term Granger causality. This is because many variables may show opposite trends in the short term, but cointegration means that they all converge in the end.

Reconsidering the VAR in Equation (1), it can be rewritten as a vector error correction model (VECM):

$${\mathsf{\Delta }\mathit{Y}\_}_{\mathit{t}}=\prod {\mathit{Y}\_}_{\mathit{t}-\mathbf{1}}+{\sum }_{\mathit{i}=\mathbf{1}}^{\mathit{p}-\mathbf{1}}{\mathsf{\Gamma }}_{\_\mathbf{i}}{\mathsf{\Delta }\mathit{Y}\_}_{\mathit{t}-\mathit{i}}+{\mathit{V}\_}_{\mathit{t}}$$

(3)

With:

• ${\mathsf{\Delta }\mathit{Y}\_}_{\mathit{t}}$: represents the differenced series at time t, which is the change in the variable of interest from the previous period.
• $\prod$: is the autoregressive parameter, and Y_(t − 1) represents the variable of interest at time t − 1.
• The summation term ${\sum }_{\mathit{i}=\mathbf{1}}^{\mathit{p}-\mathbf{1}}$: represents the autoregressive component of the model. It includes p − 1 lags of the differenced series, where p is the order of the autoregressive process.
• ${\mathsf{\Gamma }}_{\_\mathit{i}}$: represents the coefficient associated with the i-th lag of the differenced series ΔY_(t − i).
• ${\mathit{V}\_}_{\mathit{t}}$: represents the error term at time t, which captures the random or unexplained component of the model.

In fact, this vector error correction model (VECM) will take the following form:

$${\mathsf{\Delta }\mathbf{G}\mathbf{D}\mathbf{P}\_}_{\mathit{t}}={\mathsf{\alpha }\_}_{\mathbf{0}}+{\mathsf{\alpha }\_}_{\mathbf{1}}{\mathsf{\Delta }\mathbf{G}\mathbf{D}\mathbf{P}\_}_{\mathit{t}-\mathbf{1}}+..+{\mathsf{\alpha }}_{\_\mathit{P}}{\mathsf{\Delta }\mathbf{G}\mathbf{D}\mathbf{P}\_}_{\mathit{t}-\mathit{P}}+{\mathsf{\beta }\_}_{\mathbf{1}}{\mathit{P}\mathit{F}\_}_{\mathit{t}-\mathbf{1}}+..+{\mathsf{\beta }\_}_{\mathit{P}}{\mathit{P}\mathit{F}\_}_{\mathit{t}-\mathit{P}}+Y{\mathsf{\epsilon }\_}_{\mathit{t}-\mathbf{1}}+U_{\_t}$$

(4)

With:

• ${\mathsf{\epsilon }}_{\mathit{t}-\mathbf{1}}$ delayed value of the error correction term (TCM).

As with VAR, the VECM procedure is as follows: the optimum lag length for the model must be determined. To accomplish this, we begin by determining the number of lags p in the VAR(p) model using the information criteria (Akaike and Schwarz).

Next, we need to test the stationarity of the residuals. If they are stationary, then there is a long-term relationship, and we can then estimate the VECM and study the significance of the coefficients.

3.2.2. Preliminary Tests

The all-study variables are annual series, from 1990 to 2020, as shown in the table below.

Table 2. Descriptive characteristics of the variables used.

	Tax Rate	Real GDP Growth Rate
Average	22.9086	3.6347
Std. Devi	2.7602	4.0805
Maximum	28.8808	12.3729
Minimum	18.4564	−7.1871
Skewness	0.1655	−0.5553
Kurtosis	1.9637	3.8498
Jarque–Bera	1.5780	2.6075
Probability	0.4543	0.2715

Source: Authors (our calculations based on Eviews 12).

summarizes the descriptive statistics of the variables used in the estimation. The average real GDP growth rate is 3.6347%, with a minimum value of −7.1871% and a maximum value of 12.3729%.

As indicated in Table 2, the mean tax pressure rate is 22.9086%, with a minimum value of 18.4564% and a maximum value of 28.8808%. The standard deviation for this variable, as well as for the real GDP growth rate, is low, reflecting minimal volatility and thereby reducing associated risks. Both variables exhibit non-zero skewness coefficients; specifically, the tax rate series shows rightward asymmetry, while the real GDP growth rate displays leftward asymmetry. Additionally, the kurtosis coefficients for both variables are positive, suggesting that their distributions are flatter compared to a normal distribution. According to Table 2, the Jarque–Bera test probability for both series exceeds the 5% significance level, indicating that there is insufficient evidence to reject the null hypothesis (H0).

Before estimating the VAR model, a stationarity test is required to check the series and avoid problems with spurious regressions. We use the augmented Dickey–Fuller (ADF) test and the Phillips and Perron (PP) test. The results of the variable stationarity test are shown in the table below.

The stationarity study, as shown in

Table 3. Stationarity test results.

Test Results at 5% Threshold
Variables	Augmented Dickey–Filler (ADF)		Philips and Perron (PP)		Stationarity
	Critical Value	t-Statistic	Critical Value	t-Statistic	Stationarity
GDP	−1.9525	−5.8266	−1.9525	−5.8351	I (1)
PF	−1.9529	−9.6682	−1.9525	−31.5905	I (1)

Source: Authors (our calculations based on Eviews 12).

, indicates that both the real GDP growth rate and the tax pressure rate are non-stationary in levels. However, after differencing, the time series become stationary, as the t-statistics (−5.8266% and −9.6682%) fall below the critical values at the 5% threshold (−1.9525% and −1.9529%), respectively. Therefore, their probability is less than 5%.

4. Results and Discussion

To run the VAR model, it is first necessary to determine the optimum number of lags. To accomplish this, we use four information criteria (Akaike (AIC), sequential modified LR test statistic (LR), final prediction error (FPE), Hannan–Quinn (HQ)). The selection principle is to retain the number of delays equal to the one that minimizes the four selection criteria. In this study, we retain the number of lags 2, i.e., we will estimate a VAR (2) second-order autoregressive model. Based on the VAR model, the estimation results are as follows.

By comparing the t-statistics with the critical value (1.96), as shown in

Table 4. Estimation of the VAR model.

Variables	DGPD			DPF
	Correlation Coefficient	Standard Deviation	t-Statistic	Correlation Coefficient	Standard Deviation	t-Statistic
DGPD (−1)	−1.2899	0.1813	−7.1132	0.1145	0.0729	1.5700
DGPD (−2)	−0.5910	0.1740	−3.3967	0.05163	0.0700	0.7378
DPF (−1)	0.3127	0.5595	0.5590	0.0260	0.2250	0.1155
DPF (−2)	0.1002	0.5821	0.1722	−0.2899	0.2341	−1.2387
C	−0.4035	0.7041	−0.5731	0.2071	0.2831	0.7315
R-squared	0.7788			0.1444

Source: Authors (our calculations based on Eviews 12).

, we can conclude that the real GDP growth rate depends on its past values, meaning that each shock to this variable will subsequently have an impact on it. However, the results confirm that the past values of the tax pressure have had no impact on either the growth rate or the tax pressure itself.

To interpret the results of the VAR (2) model, we need to test its econometric robustness. According to Table 4, the inverse test of the roots of the characteristic polynomial is used to test the stationarity of the model, where all the roots must lie inside the unit circle.

We note that all the eigenvalues lie within the unit circle, as illustrated in Figure 6. Thus, the VAR (2) model is stationary, and the variables follow a normal distribution. In this respect, we can conclude that the model is valid and ready to be interpreted. Furthermore, if we examine causality in the Granger sense, we can see the following.

Indeed, the impact of tax pressure on economic growth has not been verified. This can be explained by the specific nature of the Moroccan tax system. The succession of reforms to the tax system over the 1990–2020 study period reflects the government’s determination to adapt the tax policy to the country’s economic evolution. These reforms generally follow international trends and seek to make taxation more growth friendly. However, despite the progress made, the Moroccan tax system continues to suffer from a number of constraints, including the superiority of indirect taxes over direct taxes, due to the size of the informal economy, fraud and tax evasion, low tax rates, exemptions for certain sectors, and inadequate control. These difficulties hamper the growth of direct tax resources, prompting the state to concentrate on indirect taxation, and consequently increasing the problem of failing to harness taxation to promote economic growth.

In this respect and based on the results of the VAR model and the causality test presented in

Table 5. Causality test.

Null Hypothesis	F-Statistic	Prob
DPIB does not Granger cause DPF	1.4997	0.2434
DPF does not Granger cause DPIB	0.1893	0.8288

Source: Authors (our calculations based on Eviews 12).

, we can see that several determinants can impact economic growth in Morocco, apart from tax pressure, and that the value of real GDP is influenced by its past values. Over the period 1990–2020, the Moroccan economy was affected by structural transformations and economic crises, which led to changes in the pace and level of economic growth. The 2008–2009 financial and economic crisis impacted Morocco’s economic growth for several years, with the deterioration in economic activity in Morocco’s partner countries transmitted with a time lag through four main channels: foreign demand addressed to Morocco, tourism receipts, transfers from Moroccans living abroad (MLA), and foreign direct investment (FDI).

In the same context, 2020 will be marked by an unprecedented economic recession. As a result, the rate of economic growth slowed down this year, affecting the years that followed, notably 2021 and 2022. We can therefore conclude that several factors other than fiscal pressure influence short- and long-term economic growth, and that the evolution of the real GDP growth rate depends on past values.

The VAR model assumes that the series are stationary. In this case, the series used are non-stationary. By differentiating them sufficiently, we can stationarize them. This operation has its limits, however, particularly if the variables share one or more stable long-term relationships. In this case, using the Engel–Granger cointegration test, as detailed in Table 5, we have tested the stationarity of the residuals resulting from this estimation. If they are stationary, then there is a long-term relationship; otherwise, the procedure stops here.

The analysis shows that the null hypothesis of the unit root is rejected, as detailed in

Table 6. Residual stationarity test.

	T-Statistic	Prob
Augmented Dickey–Fuller test statistic	−4.8586	0.0000

Source: Authors (our calculations based on Eviews 12).

. The residuals of a long-term relationship are stationary in nature, confirming the existence of a cointegration relationship between the growth rate of real GDP and tax pressure. However, the existence of a long-run relationship paves the way for the estimation of the error-correction model.

According to Granger’s representation theorem, any cointegrated system implies the existence of an error-correcting mechanism that prevents variables from deviating too far from their long-term equilibrium.

The coefficient associated with the force of the return to equilibrium is significantly negative (−0.3939%), as shown in

Table 7. Vector error-correction model using Granger’s approach.

Variable	Coefficient	Std. Error	t-Statistic	Prob
C	0.2013	0.1647	1.2222	0.2405
DLGPF (−1)	−0.0883	0.2313	−0.3820	0.7078
DLPF (−1)	3.4894	2.7793	1.2555	0.2285
DLPF	0.0800	2.8377	0.0282	0.9779
RESID LT (−1)	−0.3151	0.1237	−2.5470	0.0223
R-squared	0.6059
C	0.2563	0.1477	1.7350	0.0981
RESID LT (−1)	−0.3939	0.0748	−5.2630	0.0000
R-squared	0.5807

Source: Authors (our calculations based on Eviews 12).

. This indicates the presence of an error-correction mechanism. The mechanism reflects the convergence of the GDP series trajectories toward the long-term target. Specifically, following a shock, the GDP response variable returns to equilibrium at a rate of 39%, meaning that the shock is fully resolved after approximately two years and six months (1/0.3939 = 2.53).

This result was highlighted by Widmalm (2001), according to whom the share of public spending coming from tax revenues contributes to improving productivity and consequently economic growth. Keho’s empirical work also confirms these results. He examined the problem of tax levies on economic activity in Côte d’Ivoire. Using cointegration and causality tests on annual data covering the period from 1960 to 2006, the author obtained long-term evidence of a relationship between taxation and GDP. Thus, he asserted that taxation does not hinder growth in the long term and that tax revenues are positively correlated with the GDP and its components.

The results of the two approaches—the vector autoregression model (VAR) and the vector error correction model (VECM)—reveal the existence of a relationship between tax pressure and economic growth over the period studied. However, this relationship remains less significant and asserts that basing long-term growth on its past values and on the tax pressure remains insufficient, as the effect of taxation is difficult to quantify and the very nature of the dependent variable requires consideration of other determinants, including investment, savings, the tax structure and so on.

5. Conclusions

This study contributes to the literature by documenting the impact of tax pressure on long-term economic growth in Morocco, thus providing a framework for future research into this complex relationship. Using econometric models such as VAR and VECM, it highlights the differential impact of past values on the real GDP growth and of tax pressures on current growth. The confirmation of these results in relation to previous studies and national practices underlines their contextual relevance.

The implications of this study are of interest to researchers and practitioners alike. For researchers, it provides a basis for a deeper understanding of the mechanisms that create the relationship between tax pressure and long-term economic growth and paves the way for a broadening of the theory on the factors that create the relationship between tax pressure and long-term economic growth, such as investment, savings and tax structure. For practitioners, based on the results obtained, it is suggested that an optimized tax structure be adopted that encourages investment and innovation while regularly evaluating public spending to ensure its contribution to economic growth, particularly in the fields of education and infrastructure. Real implementation of the principles of the new tax policy reform is needed to combat tax evasion and avoidance, thereby increasing revenues without raising tax rates. It is also suggested to analyze tax progressivity to ensure a fair distribution of the tax pressure and to emphasize transparency and accountability in tax administration. Finally, before implementing new tax measures, it would be prudent to carry out impact studies to assess their potential consequences for economic growth and social equity.

It is important to recognize that the methodological choices made in this study, based on the use of specific econometric models such as VAR and VECM, may influence the generalization of the results to other contexts. Although the quantitative nature of the study offers analytical rigor, it may overlook certain qualitative aspects of the relationship between tax pressure and economic growth. As a result, it is necessary to be cautious when extrapolating the conclusions to other situations, considering the economic, fiscal and political specificities of each country. To overcome these limitations, a mixed approach integrating quantitative and qualitative analyses could be beneficial. By exploring the implications of the study design and methodology for the generalizability of the results, and by providing examples of additional contexts, this study can contribute to a better understanding of the underlying mechanisms and policy implications of the relationship between fiscal pressure and economic growth.

It is essential to recognize that recent global events such as the COVID-19 pandemic and geopolitical tensions are having a significant impact on national economies. These events may disrupt the long-term relationship between fiscal policy and economic growth in Morocco more than in other, similar contexts. However, this point deserves to be developed further in future research to provide a comprehensive view of the impact of macroeconomic and geopolitical challenges on the fiscal determinants of economic growth.

This study suggests promising avenues for future research. It invites further exploration of the determinants of the relationship between tax pressure and long-term economic growth, such as investment, savings and tax structure. In addition, the proposal of an econometric method for determining the optimal level of tax pressure will pave the way for further research. These recommendations provide valuable guidance for researchers wishing to explore these aspects in greater depth and for decision-makers looking for practical solutions to improve long-term economic growth in Morocco.