Growth dynamics of value-added tax revenue in Ghana

Ratneswary Rasiah; Turner, Jason J.; Ho, Yew Foo Article Growth dynamics of value-added tax revenue in Ghana Contemporary Economics Provided in Cooperation with: VIZJA University, Warsaw Suggested Citation: Ratneswary Rasiah; Turner, Jason J.; Ho, Yew Foo (2019) : Growth dynamics of value-added tax revenue in Ghana, Contemporary Economics, ISSN 2300-8814, University of Economics and Human Sciences in Warsaw, Warsaw, Vol. 13, Iss. 2, pp. 147-174, https://doi.org/10.5709/ce.1897-9254.305 This Version is available at: https://hdl.handle.net/10419/297475 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ www.ce.vizja.pl 147 This work is licensed under a Creative Commons Attribution 4.0 International License. The introduction of the value-added tax (VAT) to replace the sales tax in 1995 was one of the key policy steps undertaken by the government of Ghana (GoG) to further deepen and sustain the efficiency of Ghana’s tax system to boost tax revenue. This study uses quarterly data from 2000 to 2014 and employs dynamic ordinary least squares (DOLS) and Divisia Index approaches to examine the growth of Ghana’s VAT revenues and how this growth is affected by discretionary tax measures. On the whole, the study finds that all of the measures of VAT revenue (total VAT, domestic VAT, and import VAT) have experienced some growth. Growth in total VAT and import VAT is driven strongly by growth in the base, while that of the domestic VAT is driven by both discretionary tax measures and the tax base. Discretionary tax measures are found to have a depressive effect on both total VAT and import VAT revenue growth. 1. Background Value-added tax (VAT) has emerged as the most significant fiscal tool in both developed and developing economies. It has been described as a “breath-taking tax development” (Ebrill, Keen, & Summers, 2001) and the most significant event in the evolution of tax structures in the last half of the 20th century (Cnossen, 1991). Currently, more than 164 countries worldwide have adopted VAT. Approximately two-thirds of leastdeveloped countries have a VAT (Annacondia & van der Corput, 2012), and all of the 54 countries in Africa, except for six, levy a VAT (Crowe Horwarth International, 2016). As key strengths, VAT is believed to be broad based, avoiding cascading taxes (tax on another tax), thereby checking distortions and over taxation. It is also considered to have an inherent self-checking mechanism, hence the tendency to improve tax compliance and minimize the occurrence of tax avoidance, evasion and fraud. In view of the above qualities, VAT is said to have greater potential to raise a substantial amount of revenue even at moderately low rates and for countries that are even in their early stages of development (World Bank, 1991). It has also been argued that VAT revenue tends to be more stable over the economic cycle (Auerbach, 2010; Ebeke & Ehrhart, 2012). However, critics argue that gains from VAT are questionable, especially in sub-Saharan Africa (SSA) due partly to the existence of a large informal sector. Moreover, it is argued that in developing countries, there exists weak tax administrative capacity; hence, it is difficult to address gross abuse associated with exemptions, zero-rated goods and tax credit refunds, Growth Dynamics of Value-Added Tax Revenue in Ghana ABSTRACT H25, H26, E62 KEY WORDS: JEL Classification: value-added tax, consumption tax, tax collection and fiscal policy 1 School of Economics, University of Cape Coast, Ghana, 2 Dept. of Economics, University of Dar es Salaam, Tanzania Correspondence concerning this article should be addressed to: Francis Kwaw Andoh, School of Economics, University of Cape Coast, Ghana. E-mail: [email protected].gh Francis Kwaw Andoh1, Nehemiah E. Osoro2, Eliab Luvanda2 Primary submission: 11.05.2017 | Final acceptance: 23.07.2018 148 Francis Kwaw Andoh, Nehemiah E. Osoro, Eliab Luvanda 10.5709/ce.1897-9254.305DOI: CONTEMPORARY ECONOMICS Vol. 13 Issue 2 147-1742019 which have always been inherent in the structure and operation of VAT (Emran & Stiglitz, 2005). Furthermore, since the prices of some VAT goods can react more strongly to macroeconomic fluctuations, VAT revenue might be less stable (Fricke, Süssmuth, & Süssmuth, 2014). Studies on tax revenue growth abound. Recent ones include Sobel and Holcombe (1996) Bruce, Fox, and Tuttle (2006) for the US; Cotton (2012) for Trinidad and Tobago; Koester and Priesmeier (2012) for Germany; Kargbo and Egwaikhide (2012) for Sierra Leone; Wolswijk (2009) for the Netherlands; Belinga, Benedek, de Mooij and Norregaard (2014) for OECD countries; Fricke et al. (2014) for Latin American countries; Wawire (2003) for Kenya; Choudhry (1979) for the United States, United Kingdom, Malaysia and Kenya; and Kusi (1998) for Ghana. However, none of the studies paid detailed attention to VAT revenue growth. Twerefou, Fumey, Assibey and Asmah (2010), which is the study closest to ours, estimated the buoyancy and elasticity of a number of tax handles, including VAT for Ghana for the period of 1970-2007. The VAT was reintroduced in 1998; hence, at the time of the study, VAT had been in operation for only seven years. In light of this data limitation, the study merged sales tax and VAT into one variable. Obviously, the results could be more reflective of sales tax than VAT. Moreover, the focus of the study was to compare the tax buoyancy of the pre-economic reform era (before 1984) and the postreform era (i.e. after 1984). In view of the centrality of the VAT to Ghana’s fiscal consolidation efforts, examining its growth dynamics is imperative. Our key contribution is that we use exclusive and extended data on VAT revenues and apply the recently developed two-step (2S) regression approach of error correction model to estimate the longand short-term VAT buoyancy. We also employ the superior Divisia Index technique to isolate the elasticity coefficients from the buoyancy coefficients. The study also goes further to estimate the buoyancy and elasticity of the disaggregated VAT components (domestic VAT and import VAT) while at the same time, examining the asymmetric responses for each component. These dimensions are very useful for monitoring tax revenue performance and government fiscal planning. 1.1. VAT in Ghana The VAT was first introduced in Ghana in January 1995 (actual collection started in March 1995). However, after three and half months (June 1995), the VAT was suspended due to public protests resulting in several deaths. It was reintroduced in January 1998. The introduction of the VAT was in line with the recommendations of the then ongoing IMF’s Enhanced Structural Adjustment Facility (ESAF), which specifically emphasized a tax reform centered on a “broadbased consumption tax, notably a VAT”. This step was undertaken to encourage developing economies to move away from income taxation and its associated distortions and toward consumption taxation. Prior to the introduction of the VAT, Ghana’s indirect tax regime was principally based on sales tax, which was administered by the Customs, Excise and Preventive Service (CEPS), and a services tax, administered by the Internal Revenue Service (IRS). By design, the sales tax was levied only on goods and collected as a single-stage tax at the factory gate or at the point of importation, whereas the services tax was charged on hotels and restaurants (Hotels and Restaurants Customers Tax), advertisements (Advertisement Tax), sports betting (Betting Tax) and some entertainment (Entertainment Tax). As limitations, the sales tax excluded the retail and service sectors from the base of indirect taxation. This system undermined the buoyancy of the revenue system (Terkper, 1994a). In addition, sales tax provided exemptions for large numbers of supplies and end users, thereby distorting and eroding the revenue base of the tax system. Together, both the sales and services taxes were limited in scope and very complex because of the differentiated and multiple rates regime (which ranged between 10 percent and 15 percent), as well as the gamut of exemptions provided under the regime. Eventually, it became clear that the administration of these differentiated sales taxes encouraged tax misapplication, evasion and avoidance since these taxes could not meet the revenue targets (Terkper, 1994a; Osei & Quartey, 2005). The above issues implicitly defined the principal goal of Ghana’s VAT, particularly to address the fundamental problems of a distorted indirect tax regime, specifically the sales and services tax regimes and the bottlenecks exhibited by an ineffective administrative machinery (Terkper, 1994a). It www.ce.vizja.pl 149 Growth Dynamics of Value-Added Tax Revenue in Ghana This work is licensed under a Creative Commons Attribution 4.0 International License. was also to widen the scope of the net tax and broaden the tax base by roping in other economic activities that hitherto escaped the tax coverage. Thus, the VAT was meant to further deepen and sustain the efficiency of Ghana’s tax system to boost tax revenue generation (Osei & Quartey, 2005). Ghana operates a uniform (single) standard VAT rate regime with exports classified as zero rated. The uniform VAT rate has, however, undergone a number of changes. In 1995, when the VAT was first introduced, the rate was 17.5 percent. The rate was reduced to 10 percent in 1998 when the original VAT Act which established the VAT, was repealed and replaced by a new VAT Act 486 (Terkper, 1994b; VAT Act 546, 1998a). In 1999, i.e., a year after the reintroduction, the VAT rate was further increased from 10 percent to 12.5 percent (but took effect in May 2000). The additional 2.5 percent was earmarked for the Ghana Education Trust Fund (GETFund). GETFund is devoted primarily to tertiary education infrastructural development and scholarships for building the capacity of faculty for tertiary schooling (Prichard, 2009). In September 2004, the rate was again increased to 15 percent. The additional 2.5 percent was, however, branded as the National Health Insurance Levy (NHIL) because the proceeds were earmarked for the funding of the newly introduced National Health Insurance Scheme (NHIS). The NHIS had just been introduced nationwide to replace the cash and carry system of accessing healthcare (Prichard, 2009). Furthermore, in 2007, aVAT Amendment Act, VAT Act 734 was enacted to to introduce a VAT flat rate scheme (VFRS) (Ghana News Agency, 2007). This scheme is a collection mechanism that applies a marginal tax percentage rate as net VAT payable on the value of taxable supplies (Prichard, 2009). Under the VFRS, retailers and traders with annual sales ranging between GH¢10,000 ($2,272 using the current exchange rate of $1=GH¢4.4) and GH¢120,000 ($27,272) were made to charge a VAT flat rate of 3 percent as net VAT payable on their taxable goods. This scheme was intended to facilitate VAT collection in the informal retail distributive trade sector. The threshold for the payment of the VFRS was, however, increased to GH¢200,000 ($45,450) in 2017. In 2013, a new VAT Act not only increased the rate to 17.5 percent but also outlined a number of new discretionary measures (VAT Act 870, 2013). The key reforms in the Act 870 included an imposition of VAT on fee-based financial services and removal of telephone handsets from the list of VAT supplies exempted under the VAT Act 546. This implied that the supply and import of telephone sets, including mobile or cellular phones and satellite phones, attracted VAT (PricewaterhouseCoopers, 2013). However, in 2017, with the opposition party now in power, a new VAT (Amendment) Act, VAT Act 948 (2017), was enacted to amend the VAT Act 870 (Mensah, 2017). This amendment gave legal backing to the VFRS. Regarding the supply of goods, and it extended coverage to include wholesalers and retailers. The new act also reclassified the supply of financial services, domestic air transportation and sale of immovable property by real estate developers as exempt supplies. These services attracted VAT under the VAT Act 870 (The Parliament of the Republic of Ghana, 2017). In Ghana, the VAT is levied on both domestically produced and imported goods and services. The VAT on domestic goods and services is collected in stages on the value added to the product. The law governing VAT, the VAT Act 546 (VAT Act, 1998a) and VAT Regulations 1998 (VAT Regulations 1998, 1998b), define “value-added” as the mark-up that a producer or supplier adds to the cost price of his or her materials and supplies before selling the new or improved product or service to a customer (Ghana Revenue Authority, 2018). Registered manufacturers, wholesalers, retailers and services providers perform the collection on behalf of the Ghana Revenue Authority (GRA). Individuals and firms registered with the VAT Service can charge VAT on their supplies. VAT on taxable imports is collected at the time of importation. For VAT purposes, the time of importation is when the customs and other duties and taxes become due. Thus, import VAT is collected at the point of entry into the country. The value for charging VAT on imports is the value for customs duty plus import duties and other taxes, which might be chargeable by the CEPS. VAT registered importers are entitled to claim this VAT as input tax unless they render supplies exempt, in which case only the VAT incurred in producing taxable supplies will be reclaimable. According to GRA, the services provided by freight forwarding, shipping and clearing agents and port authorities are liable to VAT, and providers of these services are required to 150 Francis Kwaw Andoh, Nehemiah E. Osoro, Eliab Luvanda 10.5709/ce.1897-9254.305DOI: CONTEMPORARY ECONOMICS Vol. 13 Issue 2 147-1742019 register for VAT. The VAT on imports paid by a freight forwarding or shipping or clearing agent on behalf of an importer is an input tax that can be reclaimed by the importer and not the agent. This principle also applies to the VAT paid on any other services on behalf of the importer. However, when the agent is charged directly for a service, and the tax invoice is in his or her name, he or she is entitled to recover the VAT as an input tax but must also account for VAT in full on the services that he or she invoices to the importer (Ghana Revenue Authority, 2018). Ghana’s VAT is based on the destination principle, meaning that VAT is paid on all taxable supplies at the point of consumption, regardless of the country of production. Under the destination type, exports do not attract VAT, but imports do unless the item is an exempt supply. Appendix 1 provides a list of exempt goods and the year of exemption. Notable among the exempt goods are basic foodstuffs, medicines and pharmaceutical products, newspapers and books, water, electricity, public transportation, feed, seed, fertilizer, agricultural implements, petroleum products, and handicrafts. The invoice-credit VAT system of Ghana’s VAT allows registered traders to automatically offset the tax paid on inputs against the taxes on output, with the difference paid to the tax authorities. When the amount of input VAT deductible exceeds the output tax, the tax authority (in this case the GRA) pays the difference. However, the latter case rarely occurs given that the output tax is based on value added. This system applies to both import VAT and domestic VAT on the basis that the inputs were used wholly, exclusively and necessarily in the course of the business (The Parliament of of Republic of Ghana, 2017). 1.2. Dynamics of Tax Revenues in Ghana In broad terms, there are four categories of taxes in Ghana. They are taxes on income and property, taxes on domestic goods and services, international trade taxes and value-added tax. With the exception of the VAT, the other three categories have been operational in Ghana since independence. Prior to 1998 (before the reintroduction of VAT), taxes on international trade constituted the largest contribution to tax revenue in Ghana, followed by taxes on domestic goods and services and then taxes on income and property. Specifically, taxes on international trade contributed on average 38.8 percent to total tax revenue for the period of 1980-93 and 32.1 percent between 1996 and 1998. Taxes on domestic goods and services contributed on average 28.5 percent and 24.4 percent over the same period, respectively, and taxes on income and property contributed approximately 21.3 percent and 27.5 percent for the periods of 1980-93 and 1996-98, respectively (Osei & Quartey, 2005). However, the introduction of VAT introduced some twists into the relative contributions of the tax components. Income and property taxes became the major sources of tax revenue, contributing approximately 24.5 percent to total tax revenue in 1999, but their contribution increased steadily to 27.1 percent by 2002. VAT placed second, contributing approximately 21.9 percent of total tax revenue in 1999. This amount increased to 22.3 percent in 2002. International trade taxes, once the largest major component of tax revenue, placed third with 20.9 percent in 1999 and 19.3 percent in 2002. Taxes on goods and services contributed 14.9 percent of total tax revenue in 1999; but their contribution dropped to 10.5 percent in 2001 and subsequently increased to 14.0 percent in 2002 (Osei & Quartey, 2005). Over the last one and half decades, VAT has emerged as the largest contributor to total tax revenue, contributing on average approximately 28 percent of the total tax revenue -- almost twice the contribution of pay as you earn (PAYE) (14.1 percent) and approximately 17 times that of self-employed tax (1.58 percent). Import duties, which are the next largest contributor after VAT, account for 17.76 percent of total tax revenue, while company and PAYE contribute approximately 16.46 percent and 14.1 percent, respectively (Andoh, 2017). Figure 1 shows the contribution of VAT from 2001 to 2014. Generally, VAT revenue performance (based on VAT productivity and VAT collection efficiency) in African countries is, on average, lower than in other regions of the world, whether in the Middle East, the Western Hemisphere, Asia and the Pacific, or the European Union (WoldeMariam, 2010). Even then, Ghana’s performance relative to other lower middleincome countries is not the best (see Table 1). More than a decade into the VAT operation, Ghana is still ranked among the lowest performing countries in Africa in terms of tax-to-GDP ratio. Table 2 shows the trends in the tax-to-GDP ratio for Ghana from www.ce.vizja.pl 151 Growth Dynamics of Value-Added Tax Revenue in Ghana This work is licensed under a Creative Commons Attribution 4.0 International License. 2006 to 2015, compared to other countries in the same lower middle-income bracket. The tax-GDP ratio of Ghana has consistently remained at less than 18 percent, which is lower than the average for the developing world and even the average for Africa (Osei & Quartey, 2005; World Bank, 2014; Bank of Ghana, 2014). These developments, therefore, cast some doubt about the revenue growth performance of the much-celebrated VAT, thereby rendering a thorough analysis very crucial. COUNTRY Year VAT introduced Standard rate VAT C-efficiency VAT productivity VAT (% of tax revenue) Year of data Cape Verde 2004 15.0 0.65 0.53 37.0 2006 Morocco 1995 14.0 0.52 0.39 31.4 2006 Tunisia 1998 18.0 0.50 0.30 24.3 2009 Ghana 1998 15** 0.39 0.27 30.1 2010 Kenya 1990 16.0 0.35 0.39 28.7 2010 Cameroon 1999 19.5 0.30 0.24 37.1 2006 Cote d'Ivoire 1960 18.0 0.11 0.09 8.4 2010 Table 1. VAT Rates and Performance Among Selected Lower Middle Income Countries in Africa Source: Adapted from “Mobilizing VAT revenues in African countries” by Cnossen (2015). International Tax Public Finance, 22(6), 1077-1108. Retrieved from http://doi.org/10.1007/s10797-015-9348-1. Note: ** Ghana has revised the standard rate on four occasions: 10% in 1998, 12.5 % in May 2000; 15 % in Sept 2004; and 17.5 % in January 2015. Ranking is performed in a descending order based on VAT C efficiency Figure 1. Contribution of VAT to Tax Revenue Compared to Other Taxes Fig. 1: Contribution of VAT to Tax Revenue Compared to Other Taxes Source: Authors` own elaboration Ͳ ͷ ͳͲ ͳͷ ʹͲ ʹͷ ͵Ͳ ͵ͷ ʹͲͲͳ ʹͲͲʹ ʹͲͲ͵ ʹͲͲͶ ʹͲͲͷ ʹͲͲ͸ ʹͲͲ͹ ʹͲͲͺ ʹͲͲͻ ʹͲͳͲ ʹͲͳͳ ʹͲͳʹ ʹͲͳ͵ ʹͲͳͶ YEAR  Ǧ    152 Francis Kwaw Andoh, Nehemiah E. Osoro, Eliab Luvanda 10.5709/ce.1897-9254.305DOI: CONTEMPORARY ECONOMICS Vol. 13 Issue 2 147-1742019 2. Overview of Fiscal Development and Growth Performance Ghana’s fiscal position continues to deteriorate. Figure 2 shows Ghana’s fiscal deficit from 1996 to 2016. Although the economy registered surpluses in the mid-1980s, the deficit resurfaced in 1992 and has persisted ever since, except in 1994 and 1995 when there were surpluses of 2.8 percent and 1.7 percent of GDP, respectively (Alagidede, Baah-Boateng, & Nketiah-Amposah, 2013). The deficit decreased from 8.5 percent in 2000 to 2.4 percent in 2005. It , however, worsened, particularly over the subsequent years. It increased from 7.8 percent of GDP in 2006 to to 11.5 percent in 2008, 10.0 percent in 2013, 11.8 percent in 2014 and 10.3 percent in 2016. Ghana’s debt was consequently considered to be unsustainable, thus compelling the country to request an Extended Credit Facility (ECF) from the IMF (IMF, 2015; World Bank, 2014; Republic of Ghana, 2017). While unrestrained public expenditure is believed to have contributed to this persistent deficit, weaker nonoil tax collection is considered to be largely responsible (Amo-Yartey, 2014). The available statistics show that Ghana’s economy has experienced some steady economic growth over the past decade. Real GDP grew from approximately 3.8 percent in 1994 to 5.8 percent in 2005 and subsequently to 6.2 percent in 2006 until it peaked at approximately 15 percent in 2011 (following the commercial production of oil), thus becoming one of the fastest growing economies in the world during this year. Between 1991 and 2013, the economy’s average annual growth rate of 5.8 percent was higher relative to 3.7 percent in SSA. Per capita GDP stood at $439 in 1991; however, it increased to $502 in 2005, $930 in 2006, $1,099 in 2007 and $1,858 in 2013 (Aryeetey & Baah-Boateng, 2015; World Bank, 2014). As a result of the impressive growth performance, poverty fell from approximately 51 percent in 1991 to 39 percent in 2001 and further to 28.5 percent in 2005 (African Development Bank [AfDB], 2012; 2014). Consumption expenditures have also increased over the years. Based on Adam Smith’s ability-to-pay principle of taxation (Kendrick, 1939), therefore, these developments should be reflected in increased tax revenues. COUNTRY YEAR 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 20062015 Tunisia 24.3 25.2 26.8 26.6 27.2 29.1 29.1 29.8 30.9 30.3 27.93 Morocco 25.4 29.7 30.2 26.9 27.7 28.5 29.5 28.1 27 26.1 25.1 Cape Verde 21.1 21.3 21.7 18.7 18.8 19.9 18.2 18.1 17.4 19.2 19.44 Kenya 17.2 17.9 18.7 18 17.7 18.4 17.7 17.8 19 18.4 18.08 Africa 16.6 17.2 17.7 17.1 17.2 17.8 18.1 18.5 18.8 19.1 17.81 Cote d'Ivoire 16.2 17.1 16.9 16.8 16.5 15.3 17.7 17.6 16.9 17.6 16.86 Cameroon 14 14.8 14.8 15 14 14.8 15.3 15.8 16.1 16.4 15.1 Ghana 13.1 13.2 12.7 13.1 13.2 14.5 14.8 13.6 15.0 15.0 13.82 Table 2. Total Tax as a Percentage of GDP in Selected Lower Middle Income Countries in Africa from 2006 to 2016 Source: Adapted from “Revenue Statistics in Africa” by OECD (2016). Retrieved from http://dx.doi.org/10.1787/9789264253308-en-fr Note: Countries are ranked by authors in descending order based on the observed information in the table. Income classification is based on the World Bank classification: https://blogs.worldbank.org/opendata/new-country-classifications-2016 www.ce.vizja.pl 153 Growth Dynamics of Value-Added Tax Revenue in Ghana This work is licensed under a Creative Commons Attribution 4.0 International License. 2.1. The Concepts of Buoyancy and Elasticity Tax revenue growth performance is measured by two concepts: tax elasticity and tax buoyancy. Tax buoyancy is a measure of the total degree of responsiveness of tax revenue to both changes in the national income (national tax base) and discretionary changes in tax rates, tax rules, and tax administration. Tax elasticity, in contrast, measures the automatic response of tax revenue to changes in national income (or tax base) when the revenue effects of discretionary tax measures are removed (Bilquees, 2004; Osoro, 1993). Desirably, a good tax regime is expected to be both buoyant and elastic, i.e., yield coefficient greater than unity, indicating that the tax can potentially improve the fiscal balance through the revenue side of the budget. If changes in discretionary tax policy were revenue enhancing, the buoyancy coefficient would be greater than the elasticity coefficient. However, more preferably, the elasticity coefficient should be greater than the buoyancy coefficient to imply that the VAT regime will generate revenue automatically in response to growth in the tax base. In such a situation, tax authorities will not have to resort to frequent changes in tax policies and VAT rates to increase the revenue yield. Such practices have often led to fierce protests in Ghana. Tax buoyancy and elasticity can vary between the short and long term. Whereas the long-term responses measure the effect of economic growth on long-term fiscal sustainability, the short term captures the volatility (or the stability) of tax revenues and provides an indication of the extent to which the tax regime can be used as an automatic stabilizer. If the long-term coefficient is greater than unity, it indicates that more rapid growth of the economy (or tax base) will reduce the fiscal deficit ratio through improved revenue. Conversely, if the short-term coefficient exceeds one, it means that the tax system is highly variable, but at the same time, it could function as a good automatic stabilizer and vice versa (Belinga et al., 2014). 3. Methodology In the tax literature, the standard econometric model for the empirical estimation of tax buoyancy is obtained by regressing the log of tax revenues on the log of the relevant tax base without controls (Belinga et al., Figure 2. Ghana’s Fiscal Deficit from 1996 to 2016 Source: Adapted from “Tax reform and revenue productivity in Ghana” by Kusi (1998), Available at https://opendocs.ids. ac.uk/opendocs/handle/123456789/2210 “Annual reports” by Bank of Ghana (several series) Available at https://www.bog.gov.gh/statistics/publication/annual-report Fig. 2: Ghana’s Fiscal Deficit from 1996 to 2016  Source: Adapted from “Tax reform and revenue productivity in Ghana” by Kusi (1998), Available at https://opendocs.ids.ac.uk/opendocs/handle/123456789/2210 “Annual reports” by Bank of Ghana (several series) Available at https://www.bog.gov.gh/statistics/publication/annual-report ͹ǡ͹ ͺǡͷ ͸ǡ͵ ͸ǡͷ ͺǡͷ ͹ǡͳ ͷǡ͵ ͵ǡͶ ͵ǡʹ ʹǡͶ ͹ǡͺ ͺǡͳ ͳͳǡͷ ͻǡͷ ͸ǡͺ Ͷǡ͵ ͳͳǡͷ ͳͲ ͳͳǡͺ ͸ǡͷ ͳͲǡ͵ Ͳ ʹ Ͷ ͸ ͺ ͳͲ ͳʹ ͳͶ ȋΨȌ  154 Francis Kwaw Andoh, Nehemiah E. Osoro, Eliab Luvanda 10.5709/ce.1897-9254.305DOI: CONTEMPORARY ECONOMICS Vol. 13 Issue 2 147-1742019 2014; Dudine & Jalles, 2017; Kargbo & Egwaikhide, 2012; Kusi, 1998; Mansfield, 1972; Osoro, 1993; Sobel & Holcombe, 1996; Twerefou, et al., 2010; Wawire, 2000; Wolswijk 2009). Thus, the baseline empirical model is specified in equation (1)1. ln Rln ttt   α β Y (1) where t ln R is log of the tax revenue at time t, and Y t ln is the log of tax base at time t. The parameter β is the tax buoyancy coefficient, while t µ is the stochastic error term. Since the variables are in logs, β is interpreted as elasticity or buoyancy, depending on whether tax revenue data are adjusted for changes in discretionary policy or not. However, equation (1) might suffer from specification bias because important variables that affect VAT revenues are omitted from the model. It has been argued that VAT is more likely to perform better with improved tax administration capacity. Therefore, in line with the literature, GDP per capita (a measure of economic development) is used as a proxy for tax administration capacity and efficiency (Adam, Bevan, & Chambas, 2001; Khattry & Rao, 2002; Keen & Lockwood, 2010). GDP per capita is therefore expected to be correlated positively with VAT buoyancy. The economic structure of the economy has also been implicated in many tax performance studies. As argued by Piggott and Whalley (2001) and Emran and Stiglitz (2005), VAT becomes inferior to tariffs in the presence of a large informal sector. The rationale is that the informal sector is extremely difficult to tax because almost all of the economic activities in the informal sector are in the underground economy. Therefore, an economy with a significantly large informal sector is more likely to experience poor tax performance. As in the empirical literature, the share of agriculture in GDP is used as a proxy for the size of the informal economy. Equation (1) is, therefore, modified to capture the control variables: tttt  ln RlnN αβY  (2) where t N is a vector of control variables (in this case, log of GDP per capita and agriculture share of GDP). The variables t ln R , Y t ln , and β were defined earlier. 3.1. Empirical Strategy The two-step (2S) regression approach proposed by Sobel and Holcombe (1996) and used by Bruce et al. (2006), Wolswijk (2009) and Koester and Priesmeier (2012) is adapted to estimate equation (2). The first step involves estimating a long-term model, which requires cointegrated relationships. The second step is to obtain the short-term dynamics of tax revenue responses through an error correction model (ECM). The first-step regression is estimated using the dynamic ordinary least squares (DOLS), an estimator developed by Saikkonen (1991) and Stock and Watson (1993) and appropriate for modeling long-term and cointegrated relationships. The advantage of DOLS over ordinary least squares (OLS) is that it addresses any potential endogeneity among regressors by augmenting the model with leads and lags of the first differences of the regressor (Bruce et al., 2006; Masih & Masih, 2001). Moreover, Monte Carlo evidence on DOLS proved that the estimator is superior, especially with small samples (Stock & Watson, 1993), compared to alternative long-term estimators, such as Engle and Granger (1987) and Phillips and Hansen (1990). In terms of policy, the lags and leads of the regressor are not of interest to this study. The long-run buoyancy model for VAT is specified as: ln Rl nN ,, ,, 1 j N tt tn nj tn nt jt t ngj        αβ ΔΔ XX (3) where t ln R is the log is the observed tax revenue, , Xnj ln is the log of the tax base, and , ΔXn is the first difference of the tax base; hence, ,, ΔX j tn nt j g j     shows the j lags and leads of the first difference of the tax base. Due to the smaller sample, we limit the number of lags to one and the leads to one. The term t µ is the stochastic error term, while the parameters ,tn λ and , βtn are the coefficients of the leads and lags of the first difference of the tax base and the log of the tax base, respectively. We further modify equation (3) into an error correction model (ECM) as LnR Ne ,, 1 10 1 q NP tt tj tj tn tk ttt t nj k          δη ΔΔRΔX L nR Ne ,, 1 10 1 q NP tt tj tj tn tk tt t t nj k            δη ΔΔRΔX (4) www.ce.vizja.pl 161 Growth Dynamics of Value-Added Tax Revenue in Ghana This work is licensed under a Creative Commons Attribution 4.0 International License. From column 1, the coefficient of total VAT revenue (1.018) is statistically significant at one percent and not significantly different from unity. This finding indicates that one-percent growth in both VAT base and discretionary changes cumulatively increases total VAT revenues by 1.018 percent in the long term. This coefficient is relatively smaller than that of Twerefou et al. (2010), who obtained long-term VAT/sales tax buoyancy of 1.41 for Ghana. As explained earlier, Twerefou et al. (2010) merged sales tax and VAT data, with the latter constituting only 7 of the 37 annual data series. Consequently, the results could possibly reflect more sales tax than VAT. However, the findings are in line with Dudine and Jalles (2017), who obtained a LR buoyancy coefficient of 1.075 for a goods and service tax for low income countries. Desirably, tax buoyancy is expected to be greater than unity to imply that the tax revenue is able to improve the fiscal stance of the economy through the revenue side of the budget. However, in contrast, Ghana’s total VAT does not seem to exhibit strong long-term buoyancy. A number of factors might explain this low coefficient. First, the results point to composition effects. Total VAT is made up of both domestic VAT and import VAT. As already seen in Figure 3, the latter’s contribution to total VAT revenue (64.46 percent) is, on average, almost twice that of domestic VAT (35.52 percent). Over the years, there has been an increase in the amount of imports not subject to VAT because of exemptions and zero rating. This expansion of imports not subject to VAT could be responsible for the low buoyancy coefficient. Furthermore, it has been argued that tax evasion and tax fraud are common features of tax administration in developing countries (Keen & Smith, 2007). Further, the situation could be worse for Ghana’s VAT administration, especially in the face of (1) (2) (3) (4) TVAT DVAT+DVAT++ IVAT LR Buoyancy 1.018*** 1.454*** 1.482*** 0.966*** LR Elasticity 1.88*** 0.824*** 0.852*** 2.94*** Divisia Index -0.87 0.63 0.57 -1.98 GDP Per Capita 0.922*** 1.5199*** 3.4785** 0.288*** (0.0721) (0.2304) (0.5602) (0.0904) Agric Share of GDP -0.0598*** -1.8381** -0.5260 (0.00168) (0.5602) (0.1673)** Linktest hat (p-value) 0.96(0.13)*** 1.92(0.75)*** 1.98(0.69)*** 1.22(0.15)*** Linktest hat-sq (p-val) -0.010(0.01) -0.08(0.075) -0.08(0.069) -0.02(0.015) ADF Residual (p-val) 3.100*** 2.374** 2.373** 3.798*** Observations 57 57 57 57 R-squared 0.99 0.851 0.854 0.99 Table 5. Long-run Buoyancy and Elasticity Note: Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 ADF critical values: 1% (-2.616); 5% (-1.950) + =tax base is household final consumption ++ =tax base is total consumption (Household final consumption + Government consumption of goods and services) 162 Francis Kwaw Andoh, Nehemiah E. Osoro, Eliab Luvanda 10.5709/ce.1897-9254.305DOI: CONTEMPORARY ECONOMICS Vol. 13 Issue 2 147-1742019 a large informal sector and the frequent hikes in VAT rates. Ghana’s VAT rate has increased three times since its adoption in 1998. Given that these upward adjustments have always faced public agitation, there is every reason to believe that they have provided an incentive for VAT evasion and fraud. Total VAT revenue is also found to be elastic (1.88), meaning that a one-percent increase in the VAT base automatically increases VAT revenue by 1.88 percent. This outcome is desirable because it is an indication that revenue generation is able to automatically keep pace with the growth in the tax base, which could prevent frequent changes in tax rates that can be politically and economically costly (Osoro, 1993). Furthermore, the elasticity coefficient of total VAT is found to be greater than the buoyancy coefficient, which is the case when the overall effect of discretionary tax changes is negative, indicating that growth in VAT revenue is strongly driven by growth in the tax base and that, on the whole, discretionary measures have not been very effective in driving revenue growth. The difference between the buoyancy and elasticity coefficients is -0.87, which implies that doubling the discretionary tax measures reduces total VAT revenue by approximately 0.87 percent, which supports the results obtained under the Divisia Index that discretionary tax measures cumulatively reduce total VAT revenue by 1.85 percent over the sample period. Disaggregating total VAT into domestic and import VAT (columns 2 and 3) shows that domestic VAT revenue is found to be buoyant with a coefficient of 1.45 (or 1.48 when total consumption is used as a base) but import VAT revenue is not buoyant. Its buoyancy coefficient is less than unity (i.e., 0.966). The results thus suggest that domestic VAT has higher long-term growth potential than import VAT, although the latter’s contribution to total VAT revenue (64.46 percent) is on average almost twice that of domestic VAT (35.52 percent) (as shown in Figure 3). In terms of elasticity, domestic VAT revenue is found to be positive but less than unity (i.e., 0.82). The magnitude of the coefficient suggests that, when the revenue effects of discretionary measures are removed, a one-percent increase in private consumption (or total consumption) increases domestic VAT revenue by approximately 0.8 percent. The positive sign, conversely, suggests that domestic VAT revenue growth is driven by both growth in its tax base (automatic growth) and changes in discretionary tax measures. However, the automatic growth is stronger (0.82) relative to discretionary policies (0.63). From column (4), the elasticity coefficient of import VAT revenue is not only greater than unity (i.e., 2.94) but also considerably greater than the buoyancy coefficient (0.94). Thus, when the effects of discretionary tax measures are removed, a one-percent increase in import value is able to increase import VAT revenues by 2.9 percent. Although the elasticity coefficient of import VAT appears to be relatively higher than that of the total VAT, it is desirable because it indicates that import VAT is automatically able to keep pace with import growth. The higher elasticity coefficient could be attributed to the use of full automation procedures of clearance at Ghana’s major ports and harbors. In 2003, in a bid to improve revenue collection efficiency at the ports and harbors, the Ghana Revenue Authority installed full automation systems comprised of two components, namely, the Ghana Community Network (GCNet) and the Ghana Customs Management Systems (GCMS). The GCMS provides a fully integrated computerized system for the management of customs declarations and related activities The GCNET, conversely, is the platform that enables the GCMS to share data with all of the parties engaged in the processing of customs clearances (Andoh, 2017). Studies have shown that automating the tax administration not only reduces the compliance and administration costs but also leads to increased effectiveness in revenue collection (Vasudevan, 2007). 5.4. Results from Short-run Estimations Similar to the long term, we employ the Divisia Index to isolate the short-term elasticity from buoyancy coefficients obtained from estimating equation (2). The results are presented in Table 6. Columns 1, 2, 3 and 4 report the short-run buoyancy and elasticity coefficients for total VAT, domestic VAT (when household final consumption is used as the base), domestic VAT (when total total consumption is used as the base) and import VAT, respectively. In line with expectations, specifically in column (1), the shortrun buoyancy and elasticity coefficients for total VAT are positive at one percent, suggesting that VAT revenue moves in the same direction as its respective base. www.ce.vizja.pl 163 Growth Dynamics of Value-Added Tax Revenue in Ghana This work is licensed under a Creative Commons Attribution 4.0 International License. The buoyancy coefficient for total VAT is 0.435, while the elasticity coefficient is greater than unity (1.31). An elasticity coefficient greater than 1 implies that total VAT revenue exhibits greater fluctuations over changes in its base. Twerefou et al. (2010) obtained ashort-term buoyancy coefficient of 0.42 and elasticity of 0.7. The difference could be attributed to data since their work was based primarily on sales tax. Similarly, domestic VAT revenue exhibits positive but greater fluctuations when total consumption is used as a base (both the buoyancy and elasticity coefficients are positive and greater than unity). However, narrowing the base to only household final consumption reduces the elasticity coefficient to almost unity, suggesting that domestic VAT becomes less volatile when household consumption is used as a base, reflecting the general stability of household consumption relative to government consumption. Individuals are more likely than governments to smooth consumption in response to fluctuations over the business cycle. As shown in column (4), the import VAT elasticity coefficient of 2.19 indicates that import VAT is highly variable over fluctuations in imports, given discretionary policies. All of the coefficients for the speed of adjustment are statistically significant and have the expected negative signs. The speed of adjustment indicates that 29 percent of the deviations in import VAT revenue from its long-run value are corrected for in the next period. Those of domestic VAT are 90 percent with respect to household final consumption and 96 percent when total consumption is used as the base. The difference between the long-term import VAT buoyancy coefficient (0.97) and that of the short-term coefficient (0.22) suggests that there is a significantly large divergence in the responses of import VAT between the two time periods. However, the speed of adjustment shows that approximately 29 percent of the deviations are corrected for in the next period. The negligible divergence between the longand short-term buoyancy coefficients, coupled with the high adjustment parameter of the domestic VAT revenue, suggest that reactions of domestic revenues to changes in their (1) (2) (3) (4) Total VAT Domestic VAT + Domestic VAT ++ Import VAT Buoyancy 0.4357** 1.732** 1.796** 0. 2106** (0.1351) (0.722) (0.723) (0.1021) Elasticity 1.31** 1.10** 1.23** 2.19** EC Term -0.1037** -0.902*** -0.966*** -0.2877** (0.3761) (0.186) (0.186) (0.1258) Diagnostics Jacque-Bera 0.7906 0.40519 0.3645 0.6646 LM Test 0.2336 0.1151 0.1158 0.2579 Observations 58 58 58 57 Table 6. Short-run Buoyancy and Elasticity of VAT Revenues Note: Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 + = tax base is household final consumption ++ =tax base is total consumption (Household final consumption + Government consumption of goods and services) 164 Francis Kwaw Andoh, Nehemiah E. Osoro, Eliab Luvanda 10.5709/ce.1897-9254.305DOI: CONTEMPORARY ECONOMICS Vol. 13 Issue 2 147-1742019 base exhibit no fundamental differences between the short and long term, indicating that there is no significant overshooting or undershooting, which is an indication of a rather nondynamic pattern in domestic VAT in Ghana. Evidence of growth and stability is found for almost all of the measures of VAT. Both the longand short-term elasticity coefficients for total VAT are greater than unity, suggesting that total VAT experiences growth but, at the same time, is highly variable. Asimilar observation is made for import VAT. 5.5. Asymmetric Responses The results for the asymmetric responses are obtained by estimating equation (3). They are reported in Table 7. Columns 1 and 3 report the results for total VAT and import VAT, respectively. The total VAT coefficients for the periods of expansion and contraction in the tax base are 0.471 and 0.473, respectively, while those for import VAT are 0.40 and 0.39 respectively. This finding suggests that total VAT revenue and import VAT revenue do not exhibit any significant variability in either period. Moreover, in statistical terms, the coefficients are significantly less than one, suggesting that total VAT revenue and import VAT revenue cannot serve as good automatic stabilizers in both contractionary and expansionary periods. The case is, however, different for domestic VAT revenue. The coefficients are statistically significant and are greater than 1, suggesting greater variability and hence better revenue stability during both good and bad times. 5.6. Decomposition of Elasticities We now decompose the elasticity into tax-to-GDP, base-to-GDP and tax-to-base, enriching our understanding of the critical source of the revenue growth. The decomposition is presented in Table 8. Positive tax-to-GDP and tax-to-base coefficients are observed for total VAT in both the short and long term. Specifically, in the long term, one percent growth in GDP leads to 1.11 percent growth in total VAT revenues, while one percent growth in the base leads to 1.88 percent growth in total VAT revenues. The base-to-GDP is, however, inelastic, indicating weaker responsiveness of the tax base to growth in GDP (variation exists among components). The estimates for total VAT differ from those of Twerefou et (1) (2) (3) TVAT Domestic VAT+Import VAT Period of Expansion 0.471** 1.371** 0.403*** (0.149) (0.697) (0.127) Period of Contraction 0.473** 1.26* 0.394*** (0.148) (0.691) (0.125) EC Term (Speed of Adjustment) -0.218** -0.274*** -0.439*** (0.057) (0.046) (0.117) Jacque-Bera Test 0.255 0.3646 0.579 LM Test 0.464 0.5656 0.586 Observations 58 58 57 R-squared 0.662 0.557 0.566 Table 7. Short-run Asymmetric Responses of Value Added Tax Note: Standard errors are in parentheses *** p<0.01, ** p<0.05, * p<0.1 + denotes that tax base is private consumption www.ce.vizja.pl 165 Growth Dynamics of Value-Added Tax Revenue in Ghana This work is licensed under a Creative Commons Attribution 4.0 International License. al. (2010), who obtained a base-to-GDP elasticity coefficient of 0.79 before tax reforms and 1.11 after tax reforms in 1983. The study further obtained tax-tobase coefficients of 0.72 and 1.31 before and after the tax reform, respectively. As already pointed out, these figures reflect more sales than VAT. All three decomposed coefficients for domestic VAT estimates are less than unity in the long term. The inelastic base-to-GDP elasticity suggests that the growth in GDP is unable to induce strong growth in the base. The coefficient of tax-to-base is also inelastic and, at the same time, smaller than that of base-to-GDP elasticity, suggesting a weaker revenue-to-base response than base-to-GDP response. This observation largely reflects the problem of tax evasion and avoidance. With the presence of large informality in Ghana, it is quite easy to evade and avoid domestic VAT. The tax-to-base elasticity coefficients for import VAT are positive and greater than unity in both the long and short term. However, the base-to-GDP coefficients for both time periods are less than unity although positive. This finding suggests that the key setback to import VAT revenue growth lies in the low responsiveness of the base to growth in GDP. 6. Conclusions and Recommendations In terms of a summary, the study finds that, although import VAT dominates total VAT in Ghana, its dominance falls over time as domestic VAT’s contribution gains momentum. Moreover, both the Divisia computation and the regression analysis showed that all three VAT measures used in the study have experienced growth. Furthermore, all of the discretionary tax measures undertaken over the period have cumulatively reduced the growth of total VAT and import VAT but rather have enhanced the growth of domestic VAT revenues. Thus, while growth in domestic VAT revenue is driven by both discretionary tax measures and growth in the tax base, growth in total VAT and import VAT revenues are driven solely by growth in the respective bases. The elasticity coefficients from the error correction model show that revenues from all of the VAT measures vary positively with their respective bases but, at the same time, exhibit high variability in the short term. On the whole, a critical observation emerging from this study is that, while the key problem of domestic VAT lies in the inability of growth in the base to induce stronger automatic growth in tax Taxes Tax-to-GDP Base-to-GDP Tax-to-Base Long-run Domestic VAT+-0.82 0.82 Domestic VAT++ 0.82 0.89 0.85 Import VAT 2.23 0.86 2.94 Total VAT+++ 1.11 0.95 1.88 Short-run Domestic VAT+-1.17 1.10 Domestic VAT++ 0.91 1.25 1.23 Import VAT 2.22 0.42 2.19 Total VAT+++ 1.35 0.87 1.31 Table 8. Decomposition of Tax Elasticities Note: + =tax base is household final consumption ++ =tax base is total consumption (Household final consumption plus Government consumption of goods and services) +++= tax base is total consumption plus total imports - denotes coefficient is as obtained from domestic VAT++ 166 Francis Kwaw Andoh, Nehemiah E. Osoro, Eliab Luvanda 10.5709/ce.1897-9254.305DOI: CONTEMPORARY ECONOMICS Vol. 13 Issue 2 147-1742019 revenues, that of import VAT revenues is the inability of the base to grow automatically in response to GDP growth. This finding suggests that there cannot be aone-size-fits-all approach to boosting VAT revenues in Ghana. The approach to improving VAT revenue should be multifaceted and specific; otherwise, the revenue potential of VAT will remain fragile. The empirical findings have of important implications for Ghana’s VAT regime and therefore the fiscal system. First, the findings suggest that, in the long term, total VAT revenue has been more elastic than the erstwhile sales tax. However, in terms of buoyancy, its coefficient has been relatively smaller. For example, compared with the two previous studies of sales tax, our study obtains long-term total VAT elasticity and buoyancy coefficients of 1.88 and 1.018, respectively. Kusi (1998) obtained 1.7 and 1.58, respectively, while Twerefou et al. (2007), obtained 1.09 and 1.41. Thus, in comparative terms, total VAT generally appears to possess a better ability (relative to the sales tax) to automatically maintain pace with the growth in the tax base. However, lower buoyancy implies that VAT (relative to sales tax) was greatly damaged by the type of discretionary policies implemented during the sample period. Second, the positive and higher (greater than unity) elasticity coefficients suggest that, desirably, the VAT revenues grow automatically with growth in the base, ceteris paribus. This outcome underscores that growth in the appropriate tax base is critical for growth in VAT revenue. In this regard, this study recommends that greater emphasis must be placed on improving the respective tax bases. As shown by the findings, the informal economy (proxied by agriculture share of GDP) has a negative effect on VAT revenue. Therefore, formalizing the informal sector is believed to have agreat impact in broadening the tax base. The current government has initiated a number of measures toward digitizing the entire economy. The specific measures include the provision of national biometric identification cards (Ghana card) for all citizens by the end of 2018 and a mandatory taxpayer identification number (TIN) for all residents by end of April 2018. We recommend that these policies be pursued relentlessly since they could go a long way toward reducing the size of the informal sector and the associated level of tax evasion. Finally, the empirical results show that discretionary tax measures that were in place during the study period have cumulatively negative effects on import VAT revenue. In light of the above findings, we recommend that tax authorities should take a critical look at import exemptions regimes. A major concern is that tax exemption provisions in Ghana are not always clear and hence can easily be abused. Although there might be justifiable reasons for providing import VAT exemptions, such exemptions should be provided sparingly and should also be monitored to eliminate abuses. Such exemptions have contributed to the erosion of the import VAT base, creating the situation where, although import VAT revenue responds strongly to inbuilt growth of the base, the response of the base to growth in the economy is very weak. References Adam, C. S., Bevan, D. L., & Chambas, G. (2001). Exchange rate regimes and revenue performance in Sub-Saharan Africa. Journal of Development Economics, 64(1), 173–213. AfDB. (2014). African Economic Outlook: Promoting Youth Employment. Retrieved from https://www. undp.org/content/dam/rba/docs/Reports/African%20Economic%20Outlook%202012%20En.pdf AfDB. (2012). African Economic Outlook. Retrieved from https://www.undp.org/content/dam/rba/docs/ Reports/African%20Economic%20Outlook%20 2012%20En.pdf Alagidede, P., Baah-Boateng, W., & Nketiah-Amposah, E. (2013). The Ghanaian economy: An overview. Ghanaian Journal of Economics, 1, 4–34. Amo-Yartey, C. (2014). Improving Fiscal Management in Ghana: The Role of Fiscal Policy Rules. Legislative Alert, 21(1), 1-8. Retrieved from http://ieagh. org/wp-content/uploads/2014/08/IEA-Legislative-Improving-Fiscal.pdf Andersen, S. P. (1973). Built-in Flexibility of Sensitivity of the Personal Income Tax in Denmark. Swedish Journal of Economics, 75, 23–38. Andoh, F. K. (2017). Taxable Capacity and Efforts of Ghana’s Value-Added Tax. African Review of Economics and Finance, 9(2), 255–284. Annacondia, F., & van der Corput, W. (2012). Overview of General Turnover Taxes and Tax Rates www.ce.vizja.pl 167 Growth Dynamics of Value-Added Tax Revenue in Ghana This work is licensed under a Creative Commons Attribution 4.0 International License. March/ April. International VAT Monitor, March/ Apri(2), 2–10. Aryeetey, E., & Baah-Boateng, W. (2015, December). Understanding Ghana’s growth success story and job creation (Working Paper No. 2015/140). World Institute for Development Economics Research. Retrieved from https://www.wider.unu. edu/sites/default/files/wp2015-140.pdf Auerbach, A. J. (2010). California’s Future Tax System. California Journal of Politics and Policy, 2(3). https://doi.org/10.5070/P2W304 Bank of Ghana. (2014). Annual Report : 2014. Accra: Bank of Ghana. Retrieved from https://www.bog. gov.gh/privatecontent/Publications/Annual_Reports/bank%20of%20ghana%20annual%20report_press%202014b.pdf Belinga, V., Benedek, D., de Mooij, R., & Norregaard, J. (2014). Tax Buoyancy in OECD Countries (Working Paper No. 14/110). International Monetary Fund. Retrieved from https://www.imf.org/external/pubs/ft/wp/2014/wp14110.pdf Bilquees, F. (2004). Elasticity and Buoyancy of the Tax System in Pakistan. The Pakistan Development Review, 43(1), 73–93. Bruce, D., Fox, W. F., & Tuttle, M. (2006). Tax Base Elasticities: A Multistate Analysis of Long Run and Short Run Dynamics. Southern Economic Journal, 73(2), 315–341. Chelliah, R. J., & Sheetal, C. K. (1974). A Note on Techniques of Adjusting Tax Revenue Series for Discretionary Changes [Unpublished paper]. International Monetary Fund. Choudhry, N. N. (1979). Measuring the Elasticity of Tax Revenue: A Divisia Index Approach. IMF Staff Papers, 26(1), 87-122. Cnossen, S. (1991). Design of the Value Added Tax: Lessons from Experience. In J. Khalilzadeh-Shirazi & A. Shah (Eds.), Tax Policy in Developing Countries (p. 72-85). Washington, DC: The World Bank. Cnossen, S. (2015). Mobilizing VAT revenues in African countries. Int Tax Public Finance, 22(6), 10771108. http://doi.org/10.1007/s10797-015-9348-1 Cotton, J. J. (2012). The Buoyancy and Elasticity of NonOil Tax Revenues in Trinidad and Tobago (19902009) (Working Paper No. 06/2012).Central Bank of Trinidad and Tobago. Retrieved from https:// www.central-bank.org.tt/sites/default/files/pagefile-uploads/WP%20Buoyancy%20and%20Elasticity%20of%20Non-Oil%20Tax%20Revenues%20 in%20Trinidad%20and%20Tobago.pdf Crowe Horwarth International. (2016). Africa VAT/ GST Guide 2016. Available at https://www.crowe. com/fr/ficorec/-/media/Crowe/Firms/Europe/ fr/Ficorec/files/Africa-VAT-Guide-20161-1. ashx?la=fr-FR&hash=BB4E9A1BCBDF90BA160 AC398799E46C77376D64A Dudine, P., & Jalles, J. T. (2017). How Buoyant is the Tax System? New Evidence from a Large Heterogeneous Panel (Working Paper No. 17/4). International Monetary Fund. Retrieved from https://www. imf.org/en/Publications/WP/Issues/2017/01/20/ How-Buoyant-is-the-Tax-System-New-Evidencefrom-a-Large-Heterogeneous-Panel-44551 Ebeke, C., & Ehrhart, H. (2012). Tax Revenue Instability in Sub-Saharan Africa: Consequences and Remedies. Journal of African Economies, 21(1), 1–27. Ebrill, L. P., Keen, M., & Summers, V. P. (2001). The Modern VAT. Washington, DC: International Monetary Fund. Emran, M. S., & Stiglitz, J. E. (2005). On selective indirect tax reform in developing countries. Journal of Public Economics, 89.(4), 599–623. Engle, R. F., & Granger, C. W. J. (1987). Co-integration and Error Correction: Representation, Estimation, and Testing. Econometrica, 55(2), 251–276. Fricke, H., Süssmuth, B., & Süssmuth, B. (2014). Growth and Volatility of Tax Revenues in Latin America. World Development, 54, 114–138. Ghana News Agency (2007, July 26). VAT Flat Rate Scheme effective 1st September. Retrieved from https://www.ghanaweb.com/GhanaHomePage/ business/VAT-Flat-Rate-Scheme-effective-1stSeptember-127923 Ghana Revenue Authority. (2015). VAT exemptions. Retrieved from https://gra.gov.gh/index.php/ charging-vat/ Ghana Revenue Authority. (2018). Value Added Tax. Retrieved from https://gra.gov.gh/index.php/valueadded-tax/ Granger, C. W., & Lee, T. (1989). Investigation of production,sales and Inventory relationships using multicointegration and non-symmetric error correction models. Journal of Applied Econometrics, 4, S145–S159. 168 Francis Kwaw Andoh, Nehemiah E. Osoro, Eliab Luvanda 10.5709/ce.1897-9254.305DOI: CONTEMPORARY ECONOMICS Vol. 13 Issue 2 147-1742019 Haughton, J. (1998). Estimating Tax Buoyancy, Elasticity, and Stability (African Economic Policy Paper No. 11). Retrieved from http://pdf.usaid.gov/pdf_ docs/PNACE024.pdf International Monetary Fund. (2015). Ghana Second Review under the Extended Credit Facility Arrangement and Request for Waiver for Nonobservance of Performance Criterion -Debt Sustainability Analysis. Wahington, DC. Retrieved from https://www. imf.org/en/Publications/CR/Issues/2016/12/31/ Ghana-Second-Review-Under-the-ExtendedCredit-Facility-Arrangement-and-Request-forWaiver-43628 Kargbo, B. I. B., & Egwaikhide, F. O. (2012). Tax Elasticity in Sierra Leone: A Time Series Approach. International Journal of Economics and Financial Issues, 2(4), 432–447. Keen, M., & Lockwood, B. (2010). The value added tax: Its causes and consequences. Journal of Development Economics, 92, 138–151. Keen, M., & Smith, S. (2007). VAT Fraud and Evasion: What do we know and what can be done (Working Paper, No. 07/31). International Monetary Fund. Kendrick, S. M. (1939). The Ability-to-Pay Theory of Taxation. The American Economic Review, 29(1), 92–101. Khattry, B., & Rao, M. J. (2002). Fiscal Faux Pas?: An Analysis of the Revenue Implications of Trade Liberalization. World Development, 30(8), 1431– 1444. Koester, G. B., & Priesmeier, C. (2012). Estimating dynamic tax revenue elasticities for Germany (Working Paper No 23/2012). Frankfurt: Deutsche Bundesbank. Retrieved from https://www.econstor.eu/bitstream/10419/64816/1/726798541.pdf Kusi, N. K. (1998). Tax Reform and Revenue Productivity in Ghana (Research Paper No. 74). Nairobi: African Economic Research Consortium. Available at https://opendocs.ids.ac.uk/opendocs/handle/123456789/2210 Mansfield, C. Y. (1972). Elasticity and Buoyancy of Tax System: A Method Applied to Paraguay. Staff Papers, 19(2), 425-446. Masih, A., & Masih, R. (2001). Long and short term dynamic causal transmission amongst international stock markets. Journal of International Money and Finance, 20, 563–587. Mensah, B. (2017, March 31). Value Added Tax Amendment Bill Passed. Retrieved from http:// www.ghananewsagency.org/politics/value-addedtax-amendment-bill-passed-115081 Milwood, T.-A. T. (2011). Elasticity and Buoyancy of the Jamaican Tax System. Kingston: Bank of Jamaica. Retrieved from http://boj.org.jm/uploads/ pdf/papers_pamphlets/papers_pamphlets_Elasticity_and_Buoyancy_of_the_Jamaican_Tax_System.pdf Newey, W., & West, K. (1987). A Simple, Positive SemiDefinite Heteroskedasticity and Autocorrelation Consistent Covariance Matrix. Econometrica, 55, 703–708. Newey, W., & West, K. (1994). Automatic Lag Selection in Covariance Matrix Estimation. Review of Economic Studies, 61, 631–635. OECD. (2016). Revenue Statistics in Africa. Retrieved from http://dx.doi.org/10.1787/9789264253308-en-fr Osei, R. D., & Quartey, P. (2005). Tax Reforms in Ghana (Working Paper No. 2005/66). The World Institute for Development Economics Research. Retrieved from https://www.wider.unu.edu/sites/ default/files/rp2005-66.pdf Osoro, N. (1993). Revenue Productivity Implications of Tax Reform in Tanzania (Research Paper No. 20). African Economic Research Consortium. Retrieved from http://opendocs.ids.ac.uk/opendocs/bitstream/ handle/123456789/2037/No 20.pdf?sequence=1 Phillips, P. C. B., & Hansen, B. E. (1990). Statistical inference in instrumental variables regression with I(1) processes. Review of Economics Studies, 57(99–125). Piggott, J., & Whalley, J. (2001). VAT Base Broadening, Self Supply, and the Informal Sector. American Economic Review, 91, 1084–94. Prest, A. R. (1962). The sensitivity of yield of personal income tax in the United Kingdom. Economic Journal, 52, 576–596. PricewaterhouseCoopers. (2013, July). Amendment of Value Added Tax Act, 1998 (Act 546) as part of interim fiscal measures. Retrieved from https://www. pwc.com/gh/en/assets/pdf/ghana-on-point-vatamendment-july-2013.pdf Prichard, W. (2009). The Politics of Taxation and Implications for Accountability in Ghana 1981-2008 (Research Paper No. 330). Institute for Develop- www.ce.vizja.pl 169 Growth Dynamics of Value-Added Tax Revenue in Ghana This work is licensed under a Creative Commons Attribution 4.0 International License. ment Studies. Retrieved from https://www.ids. ac.uk/files/dmfile/wp330.pdf Republic of Ghana. (2002). Budget Statement and Economic Policy for the 2002 Financial Year. Presented by the Minister of Finance and Economic Planning: Accra, Ghana. Retrieved from https://www. mofep.gov.gh/budget-statements/2002 Republic of Ghana. (2003). Budget Statements and Finanacial Policy for 2003. Accra, Ghana: Ministry of Finance. Retrieved from https://www.mofep. gov.gh/budget-statements/2003 Republic of Ghana. (2004). 2004 National Budget Statement. Accra, Ghana. Retrieved from https:// www.mofep.gov.gh/budget-statements/2004 Republic of Ghana. (2006). Budget Statement and Economic Policy for the 2006 Financial Year. Presented by the Minister of Finance and Economic Planning : Accra, Ghana. Available at https://www. mofep.gov.gh/budget-statements/2006 Republic of Ghana. (2012). Budget Statement and Economic Policy for the 2006. Retrieved from https:// www.mofep.gov.gh/sites/default/files/budgetstatements/2012_budget.pdf Republic of Ghana. (2014). Budget Statement and Economic Policy for the 2014 financial Year. Presented by the Minister of Finance and Economic Planning : Accra, Ghana. Retrieved from https://www.mofep. gov.gh/budget-statements/2014 Republic of Ghana. (2017). Budget Statement and Economic Policy for the 2017 Financial year. Presented by the Minister of Finance and Economic Planning : Accra, Ghana. Retrieved from https://www.mofep. gov.gh/news/2017-03-02/2017-budget-statementand-economic-Policy Saikkonen, P. (1991). Asymptotically Efficient Estimation of Cointegration Regressions. Econometric Theory, 7, 1–21. Singer, N. M. (1968). The Use of Dummy Variable in Establishing the Income Elasticity of State Income Tax Revenue. National Tax Journal, 21, 200–204. Sobel, R. S., & Holcombe, R. G. (1996). Measuring the Growth and Variability of Tax Bases over the Business Cycle. National Tax Journal, 49(4), 535–52. Solow, R. M. (1957). Technical Change and the Aggregate Production Function. The Review of Economics and Statistics, 39(3), 312–320. Retrieved from http://www.jstor.org/stable/pdf/1926047.pdf Star, S., & Hall, R. (1976). An Approximate Divisia Index of Total Factor Productivity. Econometrica, 44(2), 257 – 263. Stock, J., & Watson, M. (1993). A Simple Estimator of Cointegrating Vectors in Higher Order Integrated Systems. Econometrica, 61(4), 783–820. Terkper, S. E. (1994a). Ghana: Trends in Tax Reform (1985-93). Tax Notes International, 8, 1267–1275. Terkper, S. E. (1994b). VAT in Ghana: Why it Failed. Retrieved from https://sites.hks.harvard. edu/cid/hiid/556.pdf Twerefou, D. K., Fumey, A., Assibey, E. O., & Asmah, E. E. (2010). Buoyancy and Elasticity of Tax: evidence from Ghana. Journal of Monetary and Economic Integration, 10(2), 36–70. The Parliament of the Republic of Ghana. (1998a). VAT Act 546. Retrieved from http://www.nyansaafrica.com/wp-content/uploads/2015/05/11.- The-VAT-Act-1998-Act-546.pdf The Parliament of the Republic of Ghana. (1998b). VAT Regulations 1998 (L.I. 1646). Retrieved from https://gra.gov.gh/wp-content/uploads/2018/11/ vat_regulations.pdf The Parliament of the Republic of Ghana. (2013). VAT Act 2013 (Act 870). Retrieved from https://gra.gov. gh/wp-content/uploads/2018/11/vat_act_870.pdf The Parliament of the Republic of Ghana. (2017). VAT Act 948. (Publication No. GPCL/A261/1,500/04/2017). Retrieved from http://gsiaonline.org/wp-content/ uploads/2018/07/Value-Added-Tax-AmemdmentACT-2017-Act-948-1.pdf Vacu, N. P., & Odhiambo, N. M. (2017). A Review of Imports Structure and Reforms in Ghana. Euro Economica, 36(1), 144–158. Vasudevan, R. (2007). Changed Governance or Computerized Governance: Computerized Tax Administration Processes in Tamil Nadu, India. Journal of Tax Planning and Administration, 4(1), 101–112. Wawire, N. (2000). Revenue Productivity Implications of Kenya’s Tax System. In K. Kwaa-Prah & A. G. M. Ahmed (Eds.), Africa in Transformation: Political and Economic (pp. 99–106). Addis Ababa: OSSREA. Wawire, N. (2003). Trends in Kenya’s Tax Ratios and Tax Effort Indices, and Their Implication for Future Tax Reforms. Egerton Journal, 4, 256–279. 170 Francis Kwaw Andoh, Nehemiah E. Osoro, Eliab Luvanda 10.5709/ce.1897-9254.305DOI: CONTEMPORARY ECONOMICS Vol. 13 Issue 2 147-1742019 Wawire, N. (2011, March). Determinants of Value Added Tax in Kenya. Paper presented at Centre for the Study of African Economies Conference, Oxford, UK. WoldeMariam, A. (2010). VAT revenue productivity for selected fund member countries. Washington, DC: International Monetary Fund. Wolswijk, G. (2009). The Shortand Long-Run Tax Revenue Response to Changes in Tax Bases. Economics Bulletin, 29(3), 1960–1970. Wooldridge, J. M. (2003). Introductory Econometrics: A modern approach (2nd ed.). Cincinnati, OH: Southwestern Publishing Group. World Bank. (1991). Lessons on Tax Reform. Wahington DC: World Bank Publications. World Bank. (2014). World Development Indicators. New York, NY: World Bank. WTO. (2014). Trade Policy Review. Geneva. Retrieved from https://www.wto.org/english/tratop_e/tpr_e/ s298_e.pdf Endnotes 1 In the textbook, the elasticity formula is the ratio of the percentage change in the tax revenue to the percentage change in the tax base. However, in econometric research, the formula is specified in a multiplicative form Y, t tt Ree  αβ which is then log linearized to obtain equation (1). Acknowledgments We wish to express our deep appreciation to the African Economic Research Consortium (AERC) for the financial support to carry out this research as part of the first author’s collaborative PhD programme (CPP). We are also grateful to the resource persons and members of the AERC’s thematic Group B for various comments and suggestions that helped the evolution of this study from its inception to completion. We are indebted to Prof. Nelson Wawire and also the anonymous referees who reviewed the paper and provided comments and suggestions that helped in shaping and improving the overall quality of the paper. The findings made and opinions expressed in this paper are exclusively those of the authors. The authors are also solely responsible for content and any errors.