Journal of International Development
J. Int. Dev. 30, 992–1005 (2018)
Published online 23 February 2018 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jid.3361
THE EXTENT AND DETERMINANTS OF
TAX GAP IN THE INFORMAL SECTOR:
EVIDENCE FROM GHANA
MICHAEL DANQUAH*,† and ERIC OSEI-ASSIBEY
Department of Economics, University of Ghana, Accra, Ghana
Abstract: In this paper, we attempt to estimate the tax gap in the informal sector as well as the
contributing factors of the tax losses in SSA countries using Ghana as a case study. Using micro
data on non-farm household enterprises obtained from the sixth round of the Ghana Living
Standards Survey as well as data on quarterly tax payable by specified small scale enterprises
derived from the Small Tax Payer office of the Ghana Revenue Authority, the findings show that
the national potential and actual taxes in the informal sector are US$ 81 974 846 and US$
25 023 273, respectively, reflecting an estimated national tax gap or loss of approximately US$
56 951 573. Firm level variables such as type of business, urban location and experience of the firm
significantly increase the propensity to pay tax and reduce the tax gap. Copyright © 2018 John
Wiley & Sons, Ltd.
Keywords: informal sector; taxes; tax gap; tax propensity; Ghana; sub-Saharan Africa
1 INTRODUCTION
The provision of public infrastructure and government services is a key factor for
economic development. In many developing countries, a lack of public service provision
slows down economic growth and undermines efforts to improve the living standard of
the population. There are a number of reasons for the failure of many governments in
developing countries to provide sufficient public services. A lack of tax revenue is one
of them. Taxes are the main source of public revenue and economic policies are often
based on expected tax revenue. In other words, tax policy is a fundamental component
of the economic policies of every country. In order to ensure sustained growth, it is
*Correspondence to: Michael Danquah, Department of Economics, University of Ghana, Legon, Accra, Ghana.
E-mail: mdanquah@ug.edu.gh
†The authors are grateful to the International Growth Centre, IGC, Ghana for funding the study.
Copyright © 2018 John Wiley & Sons, Ltd.
Extent and Determinants of Tax Gap in the Informal Sector 993
desirable for every government to generate tax revenue to finance essential expenditures
without recourse to excessive public sector borrowing, which often crowds out private
sector investments. The average tax revenue to GDP ratio in the developed world over
the past decade was approximately 35 per cent. In the developing countries, it was
approximately 15 per cent, whilst in sub-Saharan Africa, the average for the last decade
was 13.8 per cent. This tax gap can partly reflects weaknesses in the ability of developing
and SSA countries to raise the tax revenue required for the provision of adequate public
services.
Tax gap is commonly defined as the difference between the tax that would be raised
under a hypothetical, perfect enforcement of tax laws (potential tax) and the actual tax
payments. Tax avoidance and tax evasion are widely believed to be important factors
contributing to the tax gap (Fuest & Riedel, 2009; Ebeke, Mansour, & Rota-Graziosi,
2016; Bekoe, Danquah, & Senahey, 2016). Existing empirical studies on tax revenue
losses due to tax avoidance and tax evasion in developing countries distinguish between
a domestic component and an international component. The domestic component has
focused on tax evasion that occurs due to the domestic shadow economy or the informal
sector. Due to the limited data availability, these studies on domestic tax evasion and tax
avoidance in developing countries are typically based on macro indicators of the size of
the shadow economy or the informal sector to try to estimate the tax gap. The tax revenue
losses due to avoidance and evasion are estimated for the economy as a whole using
essentially what might be referred to as back of the envelope calculations (Aguire &
Shome, 1998; Giles, 1999; Jenkins & Kuo, 2000; Cobham, 2005). Cobham (2005) is the
most widely cited study of the domestic component of tax evasion. Cobham (2005) relies
on the shadow or informal economy estimates using data from national accounts and
macroeconomic indicators to quantify the tax gap in developing countries. Subsequently,
a number of studies on tax gap estimates for developing countries (Ahmed & Rider,
2008; Martin-Vazquez, Rider, & Wallace, 2008; Novysedlák & Palkovičová, 2012;
Harremi, 2014) have also used the macro indicator approach by Cobham (2005). The
macro indicator approach assumes that income generated in the shadow economy or the
informal sector enhances the national tax gap because this income would be taxed by
the government if perfect tax auditing was possible. Therefore, the assumptions underlying
the macro approach indicate that the informal sector or the shadow economy is the cause of
tax loss because the sector is not taxed. Therefore, some studies have also examined the
determinants of tax loss at the macro level (Yalama & Gumus, 2013; Annan, Bekoe, &
Nketiah-Amponsah, 2012; Helhel & Ahmed, 2014; Akinaboade, 2014; Antwi, Inusah, &
Hamza, 2015).
However, in sub-Saharan Africa, economic activities revolve around an ever-expanding
informal sector. The informal sector accounts for 70 per cent of employment in SSA
(AfDB, 2013). Considering the high tax evasions and avoidance within the informal sector
of many developing countries, many countries in SSA over the years have attempted to
reduce the tax gap or tax revenue losses by widening the tax net to capture players in
the informal sector or the shadowy economy in order to generate more revenue without
resorting to increasing tax rates (which has been found theoretically to depress growth).
Internal revenue regulations have therefore been amended and introduced in many SSA
countries in order to vigorously restructure and collect taxes in the informal sector (Bekoe
et al., 2016; Ebeke et al., 2016).
Many SSA countries have therefore devised different taxation tools to collect taxes from
the informal sector. Some of these taxation tools include the occupational and sector-
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
994 M. Danquah and E. Osei-Assibey
specific standards assessment; the estimated lump sum Assessment System; and the
presumptive minimum taxes. For instance, the occupational and sector-specific standards
assessment, which is a fixed lump sum tax payment to be paid by persons or enterprises
engaged in a certain business or profession; and the presumptive minimum taxes, which
is levied based on the assumed minimum income level in a given year, have been applied
in many SSA countries such as Ghana, Nigeria (in some states), Mozambique, Lesotho,
Sierra Leone, Kenya, Burkina Faso and Ethiopia among others (Taube & Tadesse, 1996;
African Tax Spotlight, 2012; Danquah & Osei-Assibey, 2016). For instance, in order to
further expand the tax base in Ghana, the Ghana Revenue Authority (GRA) introduced
the tax stamp, a presumptive minimum tax system [Internal Revenue (Amendment)
Regulations] in 2004 to capture firms in the informal sector. As indicated, all these
amendments in internal revenue regulations in SSA countries have been carried out to
ensure that the informal or shadow economy, which is assumed to represent the size of
the tax gap in the macro indicator approach, is adequately taxed. Deviating from the earlier
studies that assumes that the informal sector is not taxed, therefore represent tax loss
(macro indicator approach), this paper attempts to empirically investigate the potential
tax as well as the actual tax collected from the informal sector in order to determine the
tax gap within the informal sector in SSA. In this way, we are able to gauge how much
taxes SSA governments are collecting from the sprawling informal or shadow economy
vis-a-vis the potential tax and the tax revenue loss. In addition, we also analyse the
propensity to pay tax as well as the underlying factors explaining the tax revenue
losses/tax gap in the informal sector in SSA. Therefore, the study fills an important void
on taxation in the informal sector in SSA countries.
In this study, we use Ghana as a case study. The 2010 Population and Housing census
estimates that over 86 per cent of the Ghanaian labour force are employed within the
private informal sector, which grew by more than 6 percentage points within a decade.
The greater majority of these workers are engaged in non-farm micro/small enterprises
that are often not well organized or registered with any government agency. The latest
Ghana Living Standard Survey (GLSS 6) estimates that approximately 3.7 million
households, representing 44.3 per cent of households in the country, operate non-farm
enterprises. Following from the Internal Revenue (Amendment) Regulations, 2011, these
non-farm enterprises pay the tax stamp (a presumptive minimum tax) administered by
the Small Tax Payer office of the GRA. In this case study on Ghana, we use data on
non-farm enterprises (which includes their actual tax payments) from the sixth round
of the GLSS and data on Quarterly tax payable by specified small scale enterprises
derived from the Small Tax Payer office of the GRA to estimate the extent of the tax
gap within the informal sector and also investigate the determinants of the propensity
to pay tax as well as the underlying factors explaining the tax gap among informal
enterprises in Ghana.
The study is therefore relevant in many respects. Firstly, the findings from the study will
give an indication to policy makers about the extent of tax gap within the informal sector.
This will help to measure the prospects of mobilizing tax revenue within the informal
sector and reduce the tax gap. Secondly, the findings of the study with respect to factors
explaining tax gap in the informal sector will help policy makers to design a more effective
and appropriate policy tool to reduce the tax gap in the informal sector.
The remainder of the study is structured as follows: Section 2 provides a discussion of
the methodology and data. The findings are presented in Section 3. Section 4 focuses on
the concluding remarks.
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
Extent and Determinants of Tax Gap in the Informal Sector 995
2 METHODOLOGY AND DATA
2.1 Measurement of the Tax Gap
As indicated, the tax gap is defined as the difference between the tax that would be raised
under a hypothetical, perfect enforcement of tax laws (potential tax) and the actual tax
payments. The methods employed to quantify the (domestic) tax gap can be broadly
divided into macro approaches—which use data from national accounts and
macroeconomic indicators to quantify the tax gap—and micro approaches—which use
household or firm level data retrieved from surveys and audits. In general, micro
approaches based on information from randomly selected taxpayers are most likely to
deliver reliable tax gap estimates. In contrast, methods to quantify the domestic tax gap
based on macro indicators are less reliable and informative. Due to the shortcomings of
macro approaches, many researchers and administrators strongly prefer to rely on tax
gap estimations, which are based on micro methods. However, only a few studies have
employed the micro approach in developing countries, partly due to the lack of data and
complexity of analysis. (Slemrod & Yitzhaki, 2002; De Paula & Scheinkman, 2010;
La Porta & Shleifer, 2008).
In this case study of Ghana, we employ the micro approach using data from non-farm
household enterprises and their respective tax payments obtained from the round six of
the GLSS 6,1 which was conducted by the Ghana Statistical Service (GSS) in 2012/2013
across all the 10 regions of Ghana. The survey collected detailed information on non-farm
activities in the country including data on whether a firm pays tax or not and the actual tax
(in GH¢) paid by firms. Also, the data contains information on demographic characteristics
including gender, education and age. The informal sector in this case includes all non-farm
businesses or enterprises that are not registered with the Registrar General of Ghana. These
are mainly income generating enterprises that operate on small scale using simple skills
and are not tied to any government regulations. Given the available data on Ghana, the
sample on the informal sector in the GLSS 6 is more appropriate for the study because it
is nationally representative and a sampling weight can also be applied in order to scale
up the estimates to the national level (Younger, Osei-Assibey, & Oppong, 2017).
The activities of non-farm enterprises are generally classified into manufacturing,
trading and other activities. In Ghana, nationwide estimates show that more than two-fifths
of the households (44.3 per cent) operate a non-farm enterprise (GSS, 2013). In terms of
the gender dimension, it is observed that women operate most of the non-farm enterprises
(71.3 per cent) while their male counterparts operate only 28.7 per cent (Table 1).
In order to compute the tax gap, we first have to find the potential tax for all enterprises
in the informal sector. The non-farm dataset of the GLSS 6 does not capture the potential
tax payment by firms, therefore, we rely on the Internal Revenue (Amendment)
Regulations, 2011, which provides information on the potential tax payable by small firms
(i.e. a presumptive minimum tax called the tax stamp). The tax stamp is implemented by
the Small Tax Payer office of the GRA.2 The potential tax for each enterprise in our
1The GLSS 6 dataset is a nationally representative survey data, and hence, findings from this study can largely be
generalized.
2The Small Tax Office caters for small firms largely in the informal sector whose turn over falls below US$ 22 500
per annum and operate in permanent or temporary structures.
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
996 M. Danquah and E. Osei-Assibey
Table 1. Distribution of non-farm enterprises in GLSS 6
Frequency Percentage
Total number of households 18 000
Number of households with non-farm enterprise as a proportion of total 8519 47.0
households
Men operating non-farm enterprise 2443 28.7
Women operating non-farm enterprise 6076 71.3
Total (men and women together) 8519 100.0
Table 2. Quarterly tax payable by specified small scale enterprises
Category Rate per quarter (US$)
A
Large 11.25
Medium 7.5
Small 2.5
Table top 0.75
B
Large 8.75
Medium 5.00
Small 1.25
Table top 0.75
C
Large 6.25
Medium 3.75
Small 0.75
Table top 0.75
non-farm dataset is therefore identified by assigning the appropriate tax stamp under the
business category (Table 2) to each household firm/enterprise.
The tax stamps are assigned differently to these firms according to the size of the
enterprise and the category of activities undertaken by the firm (Table 2). Category A
activities comprises retail traders, ‘susu’ collectors,3 drinking and ‘chop bar’ owners,4
bakeries, business centres, estates and accommodation agents, block and terrazzo
manufacturers, sand and stone winners and contractors and licensed diamond and gold
winners and buyers; category B embraces dress makers and tailors, hairdressers, beautician
and barbers, artisans and hiring services, freelance photographers and car washing bays;
and category C is made up of butchers, individual undertakers, corn and other millers,
charcoal and firewood vendors, vulcanizers and alignment operators, shoe and equipment
repairs, traditional healers and other businesses determined by the minister and published
in the Gazzete.
However, the definition of the size of the enterprise (that is whether it is small, medium
or large) is not clearly spelt out in the Regulations, but this classification is required in
order to properly assign the potential tax due to each enterprise. Therefore, we devised a
working definition to help us place the firms into the three main enterprise-size type (that
3Susu collectors are a traditional form of financial intermediaries in Africa, predominantly in Ghana.
4Chop bar is a mini restaurant housed mostly in temporary structures and sells mainly local Ghanaian dishes.
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
Extent and Determinants of Tax Gap in the Informal Sector 997
is small, medium or large) so as to assign the appropriate tax stamps to the firms. Firstly,
we compute the (arithmetic) mean of the firms’ annual sales. Then, the firms that had their
annual sales less than the mean annual sales (that is, US$ 1199.75) were classified as small;
firms that had annual sales from the mean (US$ 1199.75) up to 175 per cent, that is 1.75
times the mean annual sales (thus, US$ 2099.75), were also classified as medium; and
firms that had annual sales exceeding US$ 2099.75 were defined as large. After defining
firms into small, medium and large, we then apply the appropriate tax stamp under each
business category5 (Table 2) to each household enterprise to derive the potential tax for
each household enterprise or firm. The actual tax paid by the non-farm enterprise are
recorded in the GLSS 6 dataset. The difference between the potential tax identified for each
household enterprise and the actual tax paid by the household enterprise is the tax gap for
that household enterprise. The aggregate tax gap for all household enterprises is the total
tax gap for all enterprises in our dataset.
2.2 Propensity to Pay Tax and Factors Explaining the Tax Gap
We run separate estimations to explain the firms’ propensity to pay tax as well as the tax
gap. All the estimations use the same explanatory variables (Table A2).
Model 1 Firm’s propensity to pay tax
The model explaining the firms’ propensity to pay tax is specified in Equation (1) as
follows:
PrðPT ¼ 1jXÞ ¼ Φðαþ βXÞ (1)
where the dependent variable, the probability that the firm pays tax, Pr(PT = 1), takes a
value of 1 if the firms pays tax and 0 if otherwise; Φ represent the cumulative standard
normal distribution function; α is the constant term; β is a vector of coefficients to be
estimated; andX is a vector of explanatory variables. The regressors include firm ownership,
education level of firm owners, location, savings, experience, size of firm, age and business
activity/category. Following from the intuition on the relationship between the covariates
and the propensity to pay tax, all the regressors are clearly and strictly exogenous with
the exception of the size of firm (which is constructed from the annual sales of firms). In
order to account for potential endogeneity bias with respect to the size of firm (if any),
we estimate a bivariate probit model of the propensity to pay tax. In the bivariate probit
model, we test specifically the endogeneity of the size of firm variable to check if it is
plausible that the size of firm can be reversely affected by propensity to pay tax. Table 3
presents the results of the endogeneity test. From Table 3, we find that the error terms of
the structural equations are significantly correlated, and hence size of firm is endogenous
in the tax propensity model. This indicates that without accounting for the endogeneity of
size of firm, our estimates are biased, and hence unreliable. Thus, we employ the bivariate
probit model that treats endogeneity bias with respect to the size of firm variable to examine
the factors explaining the propensity to pay tax.
5Under category A, an annual tax stamp of US$ 10, US$ 30 and US$ 45 were assigned to small, medium and large
firms, respectively. An annual tax stamp of US$ 5, US$ 20 and US$ 35 were assigned to small, medium and large
firms which are found in category B, respectively. For category C, the annual tax stamp of US$ 3 and US$ 25
were assigned to small and large firms, respectively (this category had no medium firm).
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
998 M. Danquah and E. Osei-Assibey
Table 3. Results of the endogeneity test between propensity to pay tax and size of the enterprise
(based on the bivariate probit model)
Correlation between disturbance terms Coefficient Robust std. error
Rho (ρ) 0.332* 0.188
No. of observations 7440
Notes: Rho (ρ) tests the presence of correlation between the error terms of the structural equations. If ρ is
significantly different from zero, it means that the errors are correlated, and hence there is endogeneity.
***p < 0.01,
**p < 0.05,
*p < 0.1.
Model 2 Determinants of tax gap
The determinants of the tax gap model is specified in Equation (2) as
TaxGap ¼ δþ ϕX þ ν (2)
where the dependent variable TaxGap measures the difference between firms’ potential
annual tax payment and actual annual tax payment; δ is the constant term; ϕ is a vector
of coefficients to be estimated; X is a vector of explanatory variables; and ν being the error
or disturbance term.6 The set of covariates in our tax gap model are strictly exogenous with
respect to tax gap, therefore, we do not anticipate any potential endogeneity issues. That is,
the dependent variable, tax gap does not have any effect on the set of covariates in our
model. We employ the Ordinary Least Squares technique to explain the (computed) tax
gap in the informal sector.
2.3 Definition of Explanatory Variables and the Expected Sign
Following from the objective of the study and the review of the literature on determinants
of tax evasion and avoidance, the explanatory variables employed to explain the tax gap
and propensity of firms’ to pay tax in the informal economy include firm ownership,
education level of firm owners, location, savings, experience, size of firm, age and business
category of firm. Table A2 in the appendix provides the definition of the explanatory
variables that were used in Equations (1) and (2).
3 ESTIMATION RESULTS AND DISCUSSION
3.1 The Extent of the Tax Gap
Following from the assigning of the appropriate potential tax to each household enterprise
(as indicated in Section 2.1), the potential tax revenue that can be collected from the
informal enterprises (as defined by the tax stamp) is approximately US$ 20 0485 per
annum. However, the annual actual tax collected by the GRA stood at a low figure of
US$ 51397.25, representing just about one-fourth of the potential tax. Based on these
6We test for heteroskedasticity using the White and Breusch–Pagan/Cook–Weisberg tests. After detecting the
presence of heteroskedasticity, the study used the heteroskedasticity-robust standard error approach to correct
for the problem in order to obtain robust estimates and avoid misleading results and recommendations
(Table A1).
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
Extent and Determinants of Tax Gap in the Informal Sector 999
Figure 1. Estimated national amount (in GH¢) and % for potential, actual tax and tax gap.
Source: Authors’ own computation from GLSS 6, 2012/2013. [Colour figure can be viewed at
wileyonlinelibrary.com]
statistics, the informal sector reports an annual aggregated tax gap or loss of about US$
14 9087.67, representing almost 75 per cent of the potential tax (Table A3).
The actual tax, potential tax and tax gap reported earlier were estimated using the sample
of 7440 firms. In order to scale up these estimates to the national level, we apply the
sampling weight used in the GLSS 6 data collection.7 Based on this computation, we arrive
at national potential and actual taxes of US$ 81974 846 and US$ 25023 273, respectively,
resulting in an estimated national tax gap or loss of approximately US$ 56951573
(Figure 1). In terms of Ghana’s current GDP, the national potential tax, actual tax and
tax gap represent 0.38, 0.11 and 0.26 per cent, respectively.
We proceed to provide a discussion on the disaggregation of the national tax gap
estimates based on the enterprise owner’s and enterprise characteristics (Tables A3 and A4).
We observe that, female owners are associated with a higher tax gap (US$ 122 592.55)
than their male counterparts (US$ 26 493.12). This indicates that the tax loss among female
owned firms is over four times that of male owned businesses in the economy (Table A3).
This may be explained by the fact that more women operate non-farm enterprises (71.3 per
cent) as compared with their male counterparts (28.7 per cent). With regards to the
educational level of the enterprise owner, firm owners without education have a higher
tax gap (US$ 75710.2) than those with at least primary education (US$ 73377.47). This
may imply that education has the chance of improving individuals understanding on the
importance and contribution of tax payment to government revenue, and hence, economic
development in general.
Non-farm economic activities in the rural economies account for higher tax gap or loss
(US$ 83 339.85) than the urban economies (US$ 65 747.82). Comparatively, the urban
economies are in general equipped with formalized tax administration system in terms of
logistics, skilled labour force, among others, and thereby these factors may explain the
relatively high revenue loss in the rural location.
7The command used in generating the national tax figures in STATA was tabstat variable [aw = weight], stat
(sum) format (per cent 15.0 fc), where the variable in this case represents either the sample potential tax or actual
tax figures.
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
1000 M. Danquah and E. Osei-Assibey
Considering the enterprise level characteristics and the tax gap, the incidence of tax gap
among the saving status of enterprises show that enterprises that make no savings record a
higher tax gap (US$ 82162.7) as compared with firms that make savings (US$ 66924.9).
The tax gap associated with small firms is at least five times and twice that of medium
and large firms, respectively. The magnitude of tax loss attributable to small firms may not
be unexpected because these firms are ‘too micro’ in their economic activities to be
identified, or perhaps to be ‘forced’ to be tax compliant or pay the projected or potential
tax stipulated by GRA. Another plausible explanation to this substantial tax loss could
be the fact that these small firms account for as much as 73.5 per cent of the total non-farm
enterprises in the country.
In terms of business activity and the tax gap, categories A and B that compose of ‘less
informal’ activities such as retail traders, drinking and ‘chop bar’ owners, business centres,
estates and accommodation agents, diamond and gold winners and buyers, dress makers
and tailors, hairdressers, beautician and barbers and artisans car washing bays are found
to be characterized by higher tax gap than category C.
3.2 Factors Explaining Firms’ Propensity to Pay Tax and Determinants of Tax Gap
In this section, we first discuss the results of the factors explaining the propensity to pay tax
followed by determinants of the tax gap. Tables 4 and 5 present the marginal effects
estimates of the bivariate probit and OLS estimation results for tax propensity and tax
gap, respectively. As indicated in Section 2.2, the bivariate probit is employed to control
for endogeneity of the size of firm variable.
The marginal effects estimates of the bivariate probit model indicate that male
ownership increases the propensity of tax payment relative to female ownership. These
Table 4. Marginal effects estimates of the bivariate probit model on the factors explaining the
propensity to pay tax
Bivariate probit
Variables Tax propensity (marginal effects)
Male owner 0.103** (0.0406)
At least primary education 0.174*** (0.0351)
Rural location 0.212*** (0.0336)
Firm saves 0.278*** (0.0323)
Firm’s experience 0.0166** (0.00521)
Firm’s experience square 0.000269 (0.000147)
Owners age (15–24) 0.0737 (0.0713)
Owners age (25–34) 0.00370 (0.0391)
Small firm 0.864*** (0.283)
Category A 0.224** (0.0715)
Category B 0.172** (0.0799)
Constant 0.712** (0.325)
Observations 7440
Robust standard errors in parentheses.
***p < 0.01,
**p < 0.05,
*p < 0.1.
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
Extent and Determinants of Tax Gap in the Informal Sector 1001
Table 5. OLS estimation results for the determinants of tax gap
Explanatory variables Estimated coefficients
Male owner 39.86*** (8.796)
At least primary education 15.57*** (3.823)
Rural location 27.26*** (4.891)
Firm saves 13.96** (5.064)
Firm’s experience 1.265 (1.103)
Firm’s experience square 0.0177 (0.0233)
Owner’s age (15–24) 19.68* (9.783)
Owner’s age (25–34) 19.01* (8.136)
Small firm 94.37*** (6.373)
Activity: Category A 11.84 (6.911)
Activity: Category B 18.96 (12.12)
Constant 211.5*** (14.63)
Observations 7440
R-sq 0.28
Adj. R-sq 0.26
RMSE 231.8
Robust standard errors in parentheses;
***p < 0.001.
**p < 0.01,
*p < 0.05,
results on ownership and propensity to pay tax are consistent with the findings of some
studies (McGee & Tyler, 2006; Antwi et al., 2015). In line with the findings of some
studies in the literature (Yalama & Gumus, 2013; Akinaboade, 2014 and Antwi et al.,
2015), the marginal effects estimation results for propensity to pay tax shows that having
at least primary education increases the likelihood of paying tax. The results also show that
owner’s age in the informal sector does not have any significant effect on propensity to pay
tax. Consistent with the findings of others works in the literature (Akinaboade, 2014), the
estimation results reveal that being a rural firm decreases the propensity to pay tax. Also, a
firm that makes savings significantly improves its probability of paying tax by 27 per cent
compared with its counterparts that makes no savings. The findings reveal a non-linear
effect of firm experience on propensity to pay tax. We found that at the initial stages, an
increase in firm’s experience will increase the chance of paying tax, but beyond a certain
number of years of operation, the probability of paying tax starts falling. Usually, younger
firms may tend to honour their tax obligations to avoid being closed down in the early
stages by tax official as compared with older firms who are already established or familiar
with the existing system and may sometimes hide themselves or possibly bribe corrupt
official to avoid tax payment. This finding is similar to that of Akinaboade (2014) who
posits that older firms are more likely to evade tax than younger ones. Smaller firms, that
is, firms with annual sales less than the overall mean annual sales—US$ 1199.75—are also
less likely to pay tax. The findings on type of business activity from our estimation shows
that firms in categories A and B business activities, which are mostly visible and easily
identified by tax officials, are more likely to pay tax as compared with their colleagues
in category C.
With respect to the determinants of the tax gap, the OLS estimation results show that
male ownership significantly contributes to the reduction in the tax gap by almost US$
10 (approximately GH¢40) as compared with female ownership. Also, having at least
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
1002 M. Danquah and E. Osei-Assibey
primary education lowers the tax gap by almost US$ 4. The owner’s age in the informal
sector does significantly affect the tax gap. In other words, owners in the 15–24 and
25–34 age categories tend to have a similar effect of reducing the tax gap by about US$
5 compared with those who are at least 35 years. Firms in rural localities contribute to
an increase in the tax gap by approximately US$ 7 relative to being an urban firm. It is also
established from the OLS estimation that firm’s saving habit can lower the tax gap by
about US$ 4. Firms experience has no significant effect on the tax gap. However, smaller
firms contribute significantly to the reduction in the tax gap. Firms in categories A and B
business activities, which as indicated are mostly visible and easily identified by tax
officials, contribute significantly in reducing the tax gap.
4 CONCLUSION
Using Ghana as a case study, this paper attempts to determine the extent of the potential
and actual taxes collected by governments in SSA in the informal sector. We therefore
estimate the difference between the potential tax and actual tax paid by non-farm
enterprises in the informal sector. In this way, we are able to indicate the extent of tax
revenue losses/tax gap in the informal sector. In addition, the paper also examines the
propensity to pay tax as well as the contributing factors to the tax gap in the informal
sector. In this case study of Ghana, we use micro data on non-farm household enterprises
obtained from the GLSS 6 as well as data on Quarterly tax payable by specified small scale
enterprises derived from the Small Tax Payer office of the GRA.
The findings show that the national potential and actual taxes in the informal sector are
US$ 81974 846 and US$ 25023273, respectively, reflecting an estimated national tax gap
or loss of approximately US$ 56951573. Furthermore, the empirical estimation results
indicate that several firm owner and firm level characteristics influence the propensity to
pay tax as well as the tax gap in the informal sector of Ghana. Among the firm owner’s
characteristics, evidence from the study shows that male firm ownership and having at least
primary education qualification are found to significantly increase the propensity to pay tax
and reduces the tax gap as expected. With respect to the firm-level variables, our estimates
show that firm savings, type of business, urban location as well as experience of the firm
increase the propensity to pay tax, and significantly reduce the tax gap as expected.
Although many SSA countries have large informal sector with similar characteristics,
policy recommendations from the study must be generalize with some caution because
the study focuses on Ghana. Following from the findings, it is recommended that Revenue
agencies, particularly the GRA, should intensify public education, particularly among
women micro entrepreneurs, on their tax responsibilities using different and suitable
educational platforms.
There is also the need for governments to vigorously promote financial inclusion within
the informal sector to make more entrepreneurs in the informal sector bankable as savings
in the bank was found to correlate with propensity to pay tax. New tax policies could also
incorporate innovate ways of making the payment of tax easier for rural enterprises.
Finally, as a way of motivation, the government should strengthen the expansion of
infrastructure and social facilities such as roads, hospitals and electricity for the
population to experience the benefits of taxation as a tool for their own economic
development.
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
Extent and Determinants of Tax Gap in the Informal Sector 1003
REFERENCES
African Tax Spotlight. 2012. Taxation and the informal sector. Quarterly Newsletter 2(3):
Aguire CS, Shome P. 1998. The Mexican value added tax (VAT): the methodology for calculating
the base. National Tax Journal 41(4): 543–554.
Ahmed RA, Rider M. 2008. Pakistan’s tax gap: estimates by tax calculation and methodology.
International Studies Program Working Paper no.11.
Akinaboade O. 2014. Regulation, SMEs’ growth and performance in Cameroon’s central and littoral
Provinces’ manufacturing and retail sectors. African Development Review 26(4): 597–609.
Annan B, Bekoe W, Nketiah-Amponsah E. 2012. Determinants of tax evasion in Ghana: 1970–2010.
International Journal of Economic Sciences and Applied Research 6(3): 97–121.
Antwi SK, Inusah AM, Hamza K. 2015. The effect of demographic characteristics of small and
medium entrepreneurs on tax compliance in the tamale metropolis, Ghana. International Journal
of Economics, Commerce and Management 3(3):
Bekoe W, Danquah M, Senahey KS. 2016. Tax reforms and revenue mobilization in Ghana. Journal
of Economic Studies 43(4): 522–534.
Cobham A. 2005.Tax evasion, tax avoidance, and development finance. Queen Elizabeth House
Working Paper no. 129.
Danquah M, Osei-Assibey E. 2016. Informality and the tax gap: a case of non-farm enterprises in
Ghana. International Growth Centre (IGC) Working Paper, S-33206-GHA-1. London, UK.
De Paula A, Scheinkman JA. 2010. Value-added taxes, chain effects, and informality. American
Economic Journal: Macroeconomics 2(4): 195–221.
Ebeke C, Mansour M, Rota-Graziosi G. 2016. The power to tax in sub-Saharan Africa: LTUs, VATs,
and SARAs. (no. 201611). CERDI.
Fuest C, Riedel N. 2009. Tax avoidance, tax evasion and tax expenditures in developing countries: a
review of the literature. Report for the UK Department for International Development (DFID).
Giles DEA. 1999. Modelling the hidden economy and the tax-gap in New Zealand. Empirical
Economics 24(4): 71–90.
GSS. 2013. Ghana statistical service. Ghana living standards survey-report of the sixth round.
Harrem\i M. 2014. A simple analysis of the tax gap Balkan region. Mediterranean Journal of Social
Sciences 5(19): 365–372.
Helhel Y, Ahmed Y. 2014. Factors affecting tax attitudes and tax compliance: a survey study in
Yemen. European Journal of Business and Management 6(22): 48–58.
Jenkins GP, Kuo C-Y. 2000. A VAT revenue simulation for tax reform in developing countries.
World Development 28(4): 763–774.
Martin-Vazquez J, Rider M, Wallace S. 2008. Tax reforms in Russian. Edward Elgar Publications:
Amsterdam NL.
McGee RW, Tyler M. 2006. Tax evasion and ethics: a demographic study of 33 countries. Available
at SSRN 940505.
Novysedlák V, Palkovičová J. 2012. The estimate of the value added tax revenue loss. Economic
Analysis 25: Bratislava: Institute for financial policy
Porta RL, Shleifer A. 2008. The unofficial economy and economic development (No. w14520).
National Bureau of Economic Research.
Slemrod J, Yitzhaki S. 2002. Tax avoidance, evasion, and administration. In Handbook of public
economics (Vol. 3, pp. 1423–1470). Elsevier.
Taube G, Tadesse H. 1996. Presumptive taxation in sub-Saharan Africa experiences and prospects.
Yalama GO, Gumus E. 2013. Determinants of tax evasion behaviour: empirical evidence from
survey data. International Business and Management 6(2): 15–23.
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
1004 M. Danquah and E. Osei-Assibey
Younger SD, Osei-Assibey E, Oppong F. 2017. Fiscal incidence in Ghana. Review of Development
Economics 21(4): e47–e66. https://doi.org/10.1111/rode.12299
APPENDIX
Table A1. Tests for heteroskedasticity
White’s test for heteroskedasticity
Ho: homoskedasticity against Ha: unrestricted heteroskedasticity
chi2(78) = 183.75
Prob > chi2 = 0.0000
Breusch–Pagan/Cook–Weisberg test for heteroskedasticity
Ho: Constant variance
Variables: fitted values of tax gap
chi2(1) = 3987.30
Prob > chi2 = 0.0000
Table A2. Definition of the explanatory variables used in the study
Explanatory Variable Definition
Male owner This is a dummy with a value of 1 if the firm owner is a male and 0 for female
owner
At least primary education This a dummy with a value of 1 if the owner has at least primary education and 0
otherwise
Rural location This is a dummy variable with the value of 1 if the firm is located in the rural
location and 0 for urban location
If firm does savings in the This is a dummy with the value of 1 if the firm saves part of its income and 0
bank/has bank account otherwise
Firm’s experience in years This measure the years of operation of the firm
Firm’s experience square This is the square of the experience variable. The purpose is to capture the non-
linearity effect of firm’s experience
Owner’s age This is a categorical variable for owner’s age. The age cohorts are (15–24), (25–34)
and (35 and above). For estimation purpose, each cohort represents a dummy with
the last cohort (35 and above) being the reference group
Small firm This is a dummy variable with the value of 1 if the firm is small (firms with annual
sales less than the overall mean annual sales—US$ 1199.75) and 0 otherwise
Business Activity This is a categorical variable defined into three: Categories A, B and C (see details
or definitions in the overview); Category C is set as the reference group.
Table A3. Potential tax, actual tax and tax gap by owner’s characteristics and location
Gender of owner Owner’s education Location
No
Item Overall Male Female Education Primary Rural Urban
Potential tax, actual tax and the tax gap: figures are reported in Ghana Cedis (GHC)
Potential tax 801 940 221 760 580 180 347 160 454 780 413 580 388 360
Actual tax 205 589.32 115 787.5 89 801.8 44 319.2 161 270 150588.7 55000.6
Tax gap 596 350.7 105 972.5 490 378.2 302 840.8 293 510 262 991.3 333 359.4
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
Extent and Determinants of Tax Gap in the Informal Sector 1005
Copyright © 2018 John Wiley & Sons, Ltd. J. Int. Dev. 30, 992–1005 (2018)
DOI: 10.1002/jid
Table A4. Potential tax, actual tax and tax gap by firm’s characteristics
Savings status of firm Firm size Category of activities
Item Does savings No savings Small Medium Large A B C
Potential tax, actual tax and the tax gap: figures are reported in Ghana Cedis (GHC)
Potential tax 406 700 395 240 485 320 85 500 231 120 609 040 135 200 57 700
Actual tax 139 000.4 66 588.92 113 837.9 14 948.8 76 802.6 140 201 58 593.72 6794.6
Tax gap 267 699.6 328 651.1 371 482.1 70 551.2 154 317.4 468 839 76 606.28 50 905.4