Determinants of Tax Compliance among Small Taxpayers in Western Kenya

In Kenya, tax revenues make up to 80% of the government’s budgetary resources with a negligible proportion coming from grants and loans owing to the stringent conditionalities adopted as part of Structural Adjustment programmes (SAPs) imposed on the Kenyan government in the 1990s by International monetary Fund (IMF) and the World Bank. Different taxation methods are used to create a tax base which forms a pool that the tax authority can tap thereby placing a tax burden on the populace. Kenyan tax system is, however, often characterized by increasing number of non compliance among the small taxpayers and as a result of this the government has started to focus reforms specifically tailored to the SMEs so as to improve the level of compliance in Kenya. The broad objective of the study was to investigate the determinants of tax compliance in Kakamega and Kisumu, Western Kenya. Specifically determine the effect of equity on tax compliance; establish the influence of fairness of tax regime on compliance level; establish the influence of cost of complying on compliance level and determine the effect of cost of compliance on the tax compliance level of small tax payers in Western Kenya. In order to achieve the objectives of the study descriptive design was adopted. The target population for this study was sampled from those registered with KRA as VAT a payer that is 5,000 taxpayers. Using simple random sampling a sample of 370 was obtained. The study used questionnaires to collect data. Descriptive statistics and multiple regression analysis were used to analyze data. The result indicate that only tax burden (B=.608, p=.040) and equity (B=-1.055, p=.005) were significant predictors of level of tax compliance among SMEs in Western Kenya. The other predictors: complexity (B=-.180, p=.098), Fairness (B=-.032, p=.780), cost of complying (B=.435, p=.138), and policy actions (B=.017, p=.927) were not significant predictors of level of tax compliance. The study recommended reduction of equity and increase of tax burden. Further study was required to assess the effectiveness of complexity, fairness, cost of complying, and policy actions in determining level of tax compliance.


INTRODUCTION
In Kenya, tax revenues make up to 80% of the government's budgetary resources with a negligible proportion coming from grants and loans owing to the stringent conditionalities adopted as part of Structural Adjustment programmes (SAPs) imposed on the Kenyan government in the 1990s by International monetary Fund (IMF) and the World Bank (Muriithi and Moyi, 2003). Government use taxation to raise revenues to cover their expenditures on the provision of social goods and services, to regulate the level of spending in the economy, redistribute income among the populace and control imports into the country (Wilkinson, 2007). Different taxation methods are used to create a tax base which forms a pool that the tax authority can tap thereby placing a tax burden on the populace. Kenyan tax system is, however, often characterized by increasing number of non compliance among the small taxpayers and as a result of this the government has started to focus reforms specifically tailored to the SMEs so as to improve the level of compliance in Kenya. The most critical challenge to reforms is the presence of large untaxed informal sector causing high levels of revenue leakage. It is estimated that in the financial years 2000/01and 2001/02, only 65 per cent to 69 per cent of income tax revenue was collected by KRA (KIPPRA, 2008).
The puzzle of the economic theory as to why people evade taxes has extensively been discussed since Allingham and Sandmo (1992) strongly based their theory on economic crime theory; that the extent of deterrence, as a product of being detected and the size of fines imposed determines the level of compliance. The probability of being detected is shaped by the shared knowledge among categories of taxpayers and any deterrent measure that is put in place must clearly identify the common motivations specific to that category of taxpayers (Slemrod, 2002).
Recent theories propounded by Feld and Frey (2007) argue that economic factors interact with psychological tax contract between the citizens and taxpayers to determine the compliance level among the taxpayers. Psychological tax contract is influenced by government policy, tax authority's behavior and state institutions to influence morale. Studies done by Australian Tax Office in 2004 have, however, shown that the economic consideration based on calculated risks and opportunities that would accrue from non-compliance will carry more weight for small taxpayers who are most likely to be individuals and therefore have self-interest than large taxpayers who are mostly corporate bodies. The concept of fairness in the tax structure and the contract with the government (tax authority) in explaining the compliance behavior among the taxpayers has recently started to gain currency especially after a research undertaken for OECD in 1996. The findings showed that the ability of the taxpayers to trust the regulating body and their perception of fairness with which they are treated, has a greater influence on a person's choice to contravene the tax code than simple economic self interest does (McManus and Warren, 2006).
As part of the reform measures in tax administration, the Kenya revenue Authority (KRA) was incorporated in 1995 as a semi-autonomous government agency with an amalgam of five revenue departments administering upwards of 17 revenue acts that were originally in the ministry of Finance. Since its inception, KRA has instituted various legal and administrative measures designed to seal revenue leakage through effective audit and surveillance tools and increased compliance through self assessments. Reforms initiatives that have so far been undertaken in Kenya have chiefly focused on curing shortcomings created by the complexities in the tax code, the high taxation levels and the high cost of filing returns and paying tax which clearly addresses the economic rationality inherent in the choice to comply. Whichever programme that the government adopts to improve compliance in its taxation will have a direct bearing on the incidence of tax and therefore determining the taxpayers' predisposition to comply with the tax (Allingham and Sandmo, 1992). Turnover Tax (TOT) was introduced in Kenya through an amendment to the Income Tax Act vide the Finance Bill of 2007 as one of the TMP to address the low compliance level among the small businesses in Kenya.TOT is a final tax levied at 3% of turnover and is payable quarterly on or before the 20 th of the month immediately following the quarter. In its 4 th corporate plan, noted that whereas introduction of TOT was anticipated to "reduce the administrative burden of the small taxpayers and broaden the tax base in the informal sector", it failed to achieve its 2008/2009 baseline value of 5,000 newly registered TOT taxpayers and Kshs.131 million in collections (KRA 4 th Corporate Plan) KRA (2009). Tax reforms initiatives that have so far been undertaken in Kenya have chiefly focused deterrence and curing shortcomings created by the complexities in the tax code, the high tax burden and the high cost of filing returns and paying tax without addressing the problems that are created by lack of equity and fairness in the tax system.
The introduction of TOT at the beginning of 2008 was meant to improve compliance among small businesses by reducing the tax burden, complexity and cost of complying with the existing income tax regime. KRA (2009) noted that the performance of TOT has not been satisfactory, an indication that, while it was expected to improve compliance because of its simplicity and lower rate, it fell short of having appreciable impact on the level of compliance. The failure in the conception and introduction of TOT is attributable to the limiting the possible causes of low compliance to tax burden and complexity without considering other factors that are equally at play in the determination of compliance behavior among small businesses. The broad objective of the study was to establish the determinants of tax compliance among small taxpayers in Western Kenya. Specific objectives were to: determine the effect of equity on tax compliance of small taxpayers in western Kenya; establish the influence of fairness of tax regime on compliance level of small tax payers in Western Kenya and establish the influence of cost of complying on compliance level of small tax payers in Western Kenya. The researcher was guided by the following set of questions: what is the effect of equity on tax compliance of small tax payers in Western Kenya? What is the influence of fairness of tax regime on compliance level of small tax payers in Western Kenya? And what is the effect of cost of compliance on the tax compliance level of small tax payers in Western Kenya.

Conceptual Framework
The OECD 2004 guidance notes on compliance proposes that while specific compliance obligations of taxpayers will vary from one taxation role to another, they do fall under four categories, namely: registration in the tax system; the timely filing of returns; reporting of complete and complete information in the tax returns and payment of tax obligations on time (OECD, 2004). This was shown that a risk management approach must be applied to assess the taxpayer's risk/ behaviour in order to design appropriate interventions and treatments for the non-compliance (Mahon, 1997). The compliance risk model that will best capture the appropriate interventions has two approaches to the problem of non-compliance: the first based on economic rationality while the second parameter based on behavioural issues. These independent variables help in defining critical profiling of the taxpayer behaviour and appropriate targeted strategies which will lead to better the compliance. The economic rationale and is premised on the monetary considerations on the calculated risks and opportunities that would result from going undetected. This cost benefit calculation is based on the pure rational estimation of gains and losses and severity of punishment but not the honesty or lack of it among the taxpayers (Allingham and Sandmo, 1992).
On the other hand, the behavioural issues arise of perception on the equity or lack of it in the tax system and government policies. This takes the form of the goods and services that the state provides in exchange for their tax payments. The more the government follow the benefit principle of taxation and provide public services according to the preference of taxpayers in exchange for reasonable tax price, the more taxpayers comply with the law (Feld and Frey, 2007). The Australian Tax Office (ATO) has noted in one of its findings in 2004 that the ability of the taxpayers to trust the regulating body and their perception of fairness with which they are treated has a great influence on a person's choice to contravene tax rules than simple economic or rational self interest does. In most instances when taxpayers choose to avoid their tax obligations they do not connect it with cheating the community or tax authority but is a reaction to a system that is unjust, unfair and inequitable. The economic benefits and costs of non compliance can only temper the extent to which they can act on their perceptions.

Compliance Risk Management Process
Established approaches to compliance problems in most developed countries take cognizance of the need to understand the uniqueness of their taxpayers and create appropriate tax models. Different strategies are needed to address different types of compliance behavior, and a variety of tax regimes tailored according to the risk assessment of taxpayers (McManus and Warren, 2006). The OECD 2004 guidance notes for member revenue bodies outlined compliance risk management as a series of iterative steps "for systematic identification, assessment, ranking and treatment of compliance risks" based on environmental context in order to arrive at appropriate remedies. Comparative Information Series for Tax administration in OECD countries, 2004 proposed taxpayer segment organization model which seeks to organize the policies of the revenue authority around the taxpayer's segment. Risk management help to drive the proper identification of the most critical compliance behavior based on analysis of business trends and patterns of non compliance among different categories of taxpayers.
The tax reforms initiatives that have so far been undertaken in Kenya have chiefly focused on curing shortcomings created by the complexities in the tax code, the high taxation levels and the high cost of filing returns and paying tax. In fact, the introduction of TOT in 2008 among the small taxpayers was meant to remove the complexity and costs encountered in filing income tax returns and the relatively high level of taxation. KRA however noted that "the performance of TOT has not been satisfactory" and was to embark on a review of the laws governing (KRA, 2009). This underperformance could have arisen out of the tax authority's move to adopt a tax code that was expected to allow for predictions across the whole spectrum of small businesses without investigating the underlying motivations for the non-compliance. The assumption was that the processes of tax compliance are the same for all taxpayers under all conditions. This is based on the rationale that each group of taxpayers has different characteristics and behaviours and therefore poses different risk profile to revenue.

Tax Compliance Model
There are varied reasons wherever there is high non-compliance among taxpayers but they can basically be divided into two; economic and behavioural. Slemrod (2002) argues that the assumption of risk aversion that underlies economic rationality as propounded by Allingham and Sandmo (1992) cannot explain all the compliance behaviours of taxpayers. Social and psychological deterrence proponents of tax compliance models of tax compliance argue that decision concerning tax compliance is more than just a cost-benefit calculation but other factors like morality, characteristics of the taxpayer and perceptions of fairness of tax systems among other factors play a part in understanding of tax compliance decisions (Wenzel, 2002). This models moves away from the economic self-interest that had been outlined by Allingham and Sandmo (1992) which hold that people are motivated entirely by economic welfare where they calculate all associated risks and opportunities and disobey the law when they anticipate the probability of being got are lower than the economic gain that would result from non-compliance (Allingham and Sandmo, 1992).
According to the Australian Tax Office, any model which seeks to address the reasons for non-compliance must address the economic interests of the taxpayer, the" procedural justice" in the conception and design of tax system and the "distributive effects" of the tax regime. Taxpayers are considered rational actors who want to maximize their individual outcomes. In the model, tax compliance would increase with the increase in detection probability and the severity of penalties and sanctions (Mason et al., 1984). Other studies undertaken in OECD countries in 2004 has shown strong evidence across various context that perceived fairness of procedures involved in decision making and allocation processes affect the general acceptance of the tax regime. Procedural justice is be measured by the taxpayer input in formulating tax laws and whether enforcement procedures employed are fair. People would generally desire a process and decision control because it would give them control over the outcomes. It involves asking whether taxpayers have any say in the formulation of the tax code.
Distributive justice refers to the taxpayers perception as to whether there is fairness of tax burden borne by an individual or group as compared to the burden borne by other people or groups in the same category and secondly, if there is fair exchange between the taxpayers and the government in services received for taxes paid. Perceived inequity increases tax evasion and since most SMEs are managed at the individual level, these above variables tend to be rationalized and magnified than if the business is a corporation which is theoretically considered impersonal (Wenzel, 2002). According to ATO the value of this model is in the contribution it makes in developing a deeper understanding of taxpayer behaviour and being able to lay the groundwork for the development of targeted strategies which encourage the taxpayer to do the right thing and constrain the motivation to resist or evade compliance.

Tax Reforms in Kenya
Compliance is critical in tax revenue collection in that without high compliance from taxpayers there will be low tax revenue collection by the Kenya Revenue Authority. The Kenyan tax system is often characterized by increasing number of non compliance among the small taxpayers and as a result of this the government has started to focus reforms specifically tailored to the SMEs so as to improve the level of compliance in Kenya (KIPPRA, 2008). As part of reform package in the financial management agreed between IMF, World Bank and the Kenyan government, KRA was incorporated in 1995 as an autonomous government revenue agency with three main revenue departments of customs and Excise, sales Tax and Income tax that were initially in the Ministry of Finance. According to the Institute of Economic Affairs (IEA), the most prominent challenge to reforms is the presence of large untaxed informal sector as well as high levels of revenue leakage. It is estimated that in the financial years 2000 and 2002, only 65 per cent to 69 per cent of income tax revenue was collected by KRA (KIPPRA, 2008).

Tax Rationalization
IEA has attributed poor performance of Income Tax in Kenya to numerous tax holidays, investment allowances and tax credits, accelerated depreciation, investment subsidies and indirect subsidies which have introduced inequities in the administration of tax. On the other hand, the introduction of VAT in the 1990s as an indirect tax and touted as "the tax of the future", was hampered by its broad based nature with numerous rates and complex input credit system with long verification process that created loopholes.
The reforms that have been made over the years are geared towards enhancing tax compliance by simplifying assess tax and streamlining filing of returns. Compliance surveys by KIPPRA and Ministry of Finance in 2004 indicated that 51% of taxpayers view Income Tax as high, 85.9% view Pay As You Earn (PAYE) as very high while 60% consider VAT as punitive and require paid accountants to prepare returns and make payments.
www.gjournals.org  (2009) cited the cumbersome payment processes and returns as major impediments to compliance levels and proposed a number of measures to reduce the complexity of forms and time and costs taken to file tax returns. The self assessment system was introduced for Income Tax and VAT in 1992 as a strategy to enhance tax compliance by making it easier for taxpayers to assess their own tax liability, file return, report that assessment and pay taxes due (KIPPRA, 2008). This system allowed for proper identification of all entities required to pay taxes through the issuing of unique personal identification number (PIN) that could be fed into a master file upon which updates are made and retrievals done much faster. The major drawback to revenue enhancement by self assessment system is the large share of informal sector in Kenya -accounting for 18% of gross domestic product (GDP). Many small scale thriving businesses have unlimited scope of evading tax based on their perception regarding costs associated with compliance and the economic benefits of going undetected. They are also largely illiterate or lack access to affordable expert advice on tax matters.

Compliance Costs
The requirement for taxpayers to use electronic tax registers (ETR) in Kenya was introduced through the Finance Act of 2004. The VAT Act was amended to empower the commissioner to require any person to use ETR to generate cash sale receipts (Finance Act, 2004). According to KRA (2009), the use of ETR is meant to reduce fraudulent accounting, authenticate documents, reduce paperwork and improve management of business through electronic connectivity. Numerous questions arise in this regard including how should transaction records be characterized, the appropriate record keeping format and standard, and how detailed should the description should be? This challenge assumes an even more pronounced complexity with the SME's compliance since it is widely believed that tax administration is regressive with the size of the firm, in the sense of taking a larger portion of resources for smaller companies as compared to the larger ones (Slemrod, 2002).

Introduction of Turnover Tax
Turnover Tax (TOT) is the Kenya government's latest Tax Modernization Programme (TPM) exclusively targeting the growing untaxed SMEs introduced in 2008 through an amendment to the Income Tax Act Cap 470. The tax is applicable to any resident person whose turnover from business exceeds Kshs.500,000 but does not exceed Kshs, 5million during any year of income. For income tax purposes, TOT is a final tax charged at 3% on gross sales with no expenditure or capital allowances and is paid quarterly on the 20 th day of the month immediately following end of the quarter. It has been noted that compliance costs for taxpayers in developing countries are four times higher than those in developed economies and is much more pronounced among SMEs since complexities in the tax system increases the cost of engaging tax experts to file returns and interpret the numerous tax deduction rules, offsets, exemptions and allowances. Most tax authorities have noted that the small businesses in the informal sector are many and heterogeneous rendering them "unrewarding in terms of income supplementation and unpleasant and dangerous to tax" (OECD, 2004).
The rationale of introducing TOT in 2008 was therefore based on the realization that compliance is dependent upon a highly monetized economy with high literacy among the business community capable of maintaining accurate accounting records and up-to-date and reliable data -characteristics which are lacking among the SMEs in Kenya. The burden of tax compliance is more likely to be felt by small businesses chiefly because the complexity in the tax code overwhelms the resources of the taxpayers and causes inadvertent non-compliance or errors of omission (Slemrod, 2002). With TOT, KRA (2009) hoped to achieve efficiency and effectiveness in tax administration by encouraging self assessment and compliance that would reduce compliance and administrative costs through rationalization of the tax rates pegged at 3% of turnover.

Rationale for Taxing SMEs
Whereas it has been noted that it is unrewarding and difficult to tax SMEs, tax authorities worldwide recognise the importance of compliance in this sector because reduces moral hazard among taxpayers, removes competitive distortions and help improve record keeping. When a tax authority ignores the SMEs in tax enforcement or subject them to "friendly" regime as in the case of TOT, a situation is created where other taxpayers shift from "good" compliance to bad compliance equilibrium. There is a possibility of the larger businesses' compliance being undermined when smaller ones are seen not to pay reasonable amounts. Competitive distortions and inefficiencies will arise as firms begin to have a disincentive to regularize their processes where sections of businesses are left out of tax brackets. When TOT was initially introduced by KRA in 2008, most firms rushed to register with the regimean indication that there was an artificial limitation in business growth and splitting of operations to make them fall within the TOT bracket. Finally efforts to bring SMEs under the normal tax regimes will eventually boost the record keeping capabilities and financial sophistication thereby improving capital process and government revenues.

Empirical studies
The causes of non-compliance among taxpayers ( based on filing , accuracy and payment) is for the purpose of this study treated the same as tax evasion which encompass willful attempt to evade tax liability by submitting false information and records or omitting any material details that should have been disclosed. This must be differentiated with tax avoidance which involves techniques of minimizing tax liability through legal means. Several studies have been done on the causes of tax evasion and in most instances the trade-off between the gain if the evasion is undetected and the loss if evasion is detected and penalized has been the dominant theme. This cost benefit calculation has been premised on the pure rational permutations of gains and losses and not the honesty or lack of it among the taxpayers (Allingham and Sandmo, 1992).
Firm level surveys carried in the World Competitiveness report [2004][2005] indicate that Kenyan firms are highly taxed with tax index burden of 3.9 as compared to a mean of 3.6 for 104 countries in the study. However, it should be noted that lowering the tax rate to bring the tax burden to the mean faces the risk of revenue loss if the lower tax rate does not improve compliance. Gross increase in the tax rates or high levels of taxation motivate taxpayers to find ways of reducing tax liability. Whereas large taxpayers have the expertise and resources to do tax planning thus reducing liability by legal means, the small taxpayers due to their limited in resources will invariably tend to evade tax. Allingham and Sandmo, (1992), modelled the taxpayer compliance behavior on a gamble of an individual's choice regarding tax evasion which is basically determined by a trade-off of a gain if the evasion is not detected and loss when found and penalized. Various studies done by ATO in 2004 have, however, shown that the economic consideration based on calculated risks and opportunities that would accrue from non-compliance will carry more weight for small taxpayers who are most likely to be individuals and therefore have self-interest than large taxpayers who are mostly corporate bodies (McManus and Warren, 2006).
The concept of fairness in the tax structure and the contract with the government (tax authority) in explaining the compliance behavior among the taxpayers has recently started to gain currency especially after a research undertaken for OECD in 1996. The findings showed that the ability of the taxpayers to trust the regulating body and their perception of fairness with which they are treated, has a greater influence on a person's choice to contravene the tax code than simple economic self interest does.
Kenya's tax structure has numerous pieces of legislations and regulations prescribing complex deductions and allowances, extensive bands and brackets and grafted in technical jargon as to confuse most small taxpayers. Entangling the complex tax laws also overwhelms the resources of most taxpayers causing inadvertent noncompliance (Slemrod, 2002). Most small taxpayers are quite frustrated with the convoluted tax code and will invariably resort to "creative" accounting to escape the tax net. Multiplicity also allows the taxpayer to "slide from one type of tax to another or slip from a higher to lower marginal tax rates solely to reduce tax liability". The magnitude of financial, human and time resources that taxpayers have to invest to comply with the tax laws lead to low compliance level. The cost of engaging tax experts to file returns and interpret the numerous deduction rules, offsets, exemption and allowances overwhelms the small taxpayers (Moyi and Ronge, 2006). As was noted by Slemrod (2002) in the study of compliance behaviour, the cost of compliance is fixed and will not significantly vary between the large and small taxpayers thus making it regressive with respect to firm size, in the sense of being larger in proportion to size for smaller taxpayers as compared to the larger ones.

Research Design
This study was descriptive in nature. Research design is the plan and structure of investigation so conceived to obtain answers to research questions. It gives the framework and organization of the relationships among the variables in the study (Cooper and Schindler, 2007). Descriptive research is concerned with the qualitative phenomena and this study used it to determine the factors affecting taxpayer compliance and how they impact on the level of compliance among the small taxpayers.

Study Area
The study covered taxpayers in Kisumu in Nyanza province and Kakamega in Western province of Kenya targeting the SMEs. Although there is no universal agreement on how small businesses are defined, in Kenya SMEs have been defined as those enterprises that have between 1 and 50 employees (Kenya Economic Survey, 2009). The two towns in the region were picked from Western Kenya because of close proximity to the researcher and the fact that most businesses and these locations are predominantly SMEs.

Target Population
The target population for this study was sampled from those registered with KRA as VAT payers. As at close of 2008 there were 5,000 taxpayers for VAT in Kisumu and Kakamega (KRA, 2009). The other category of respondents was all tax officers working with the Compliance and Audit Programmes in Domestic Taxes Offices (DTD) in Kakamega and Kisumu.

Sampling Technique and Sample size
This subsection covers the sample size and sampling technique. A sampling frame according to Cooper and Schindler is a list of elements from which the sample is actually drawn and is closely related to the population. The sampling frame was SME bracket taxpayers in Nyanza and Western Provinces who are registered for VAT currently numbering 5,000. A sample size was determined based on these registrants. A formula propounded by Cochran (1963) was used to determine the size as follows; Where; n -is the sample size N -is the population size e -is the level of precision(95%; e = 0.05) From the above a sample of 370 respondents was picked.

Data Collection
Data was collected from the taxpayers identified from KRA registered taxpayers in Kisumu and Kakamega. The study obtained primary data through the use of questionnaires. Questionnaires were handed to the taxpayers by the research assistants and requested to fill or be assisted to fill them. The questionnaires were structured for ease of administering and analysis. The study used structured questionnaires to collect primary data and was administered to the 100 potential respondents in Kisumu and Kakamega.

Validity of Instruments
Validity involves the degree to which obtained results from data analysis actually represents the phenomenon under study. Validity is tested by representativeness of the target population and by consensual judgements by experts (Mugenda and Mugenda, 1999). During the pilot study, the researcher focused on objectives of the study and diligently determined the parameters to be measured in order to establish the existing relationship among the variables.

Reliability Test
The results from the first trial were recorded and after the second trial the results will then be compared and correlated with the initial test to give a measure of reliability while taking into account the time differences. The two week period is a reasonable period between the two testing sessions since there was minimal time variation. Validity involves the degree to which obtained results from data analysis actually represents the phenomenon under study. During the pilot study, the researcher focused on objectives of the study and diligently determined the parameters to be measured in order to establish the existing relationship among the variables. Questionnaire was constructed using simple language so that the respondents are in a position to answer questions with least misrepresentations.

Data Analysis
The study examined the collected data to make inferences through a series of operations involving editing to eliminate repetitions and inconsistencies, classification on the basis of response homogeneity and subsequent tabulation for the purpose of inter-relating the variables under study. The data was analyzed using descriptive statistics and multiple regression analysis of the form: . ...

Socio demographic Characteristics
Of the 370 respondents, 206 respondents returned the questionnaires. Thus there was a response rate of 55.67% this was attributed to the respondents fearing that it might have been a trap for checking their tax compliance status. From Table 2 below, 54.4% of the respondents had operated the business for a period between 3-5 years, 4.4% 5-7 years, 15.7% 1-4 years, and 24.85% less than 1 year.    From Table 4, the mean duration of filled tax returns in the last 5 years is 1.5631, mean value for the degree of complexity is 2.1699,fairness is 3.4466,cost of complying 3.2621,tax burden 3.2621,equity is 1.4272 and level of compliance is 1.4 as Table 4 indicates.

Multiple Regression Analysis
In order to measure the determinants of the tax compliance multiple regression Eq. 1 was estimated. The summary of the model is displayed in Table 5. The R-squared value for the model indicates that the explanatory variables accounts for 80% of the variations in the dependent variable (tax compliance).The results also indicate that the model is stable and significant at both 1% and 5% significance level ( see Table 6). With Durbin-Watson value of 2.001 in table 5 above it indicates that the regression model does not suffer from autocorrelation.

DISCUSSIONS
With the current results indicating public policy not being significant, the results do not concur with the Everst-Philips and Sandall (2009) who while examining the relationship between public governance quality and compliance noted that there is a linkage between public governance quality and taxation and that quality governance deliver good tax system and equally better tax system make it possible to have good governance. Akpo (2009) equally stated that good governance entails the provision of quality public goods to the public and that where government fails to provide public amenities and infrastructure to the citizen in exchange for tax payment, citizen may become reluctant to pay tax. The study of Alm and Gomez (2008) established significant positive association between perception of the benefits to be derived from political goods and the willingness of taxpayers to comply with tax laws. The same applies to fairness the findings, our results do not concur with the results by Moser, Evans III and Kim (1995) who seem to suggest that perceived fairness of the tax rate is more important than its absolute level. Judging the fairness of taxation, however, requires comprehensive knowledge and correct interpretation of the tax law, and complex tax rate structures, such as progressive taxation, that are not perfectly understood by taxpayers.
As indicated in the current study complexity, cost of complying equity does not significantly affect level of tax compliance. In the early stages of the introduction of SAS, taxpayers felt that the tax system was too complex (Loo, 2006) and this discouraged taxpayers from filing tax returns voluntarily. They also perceived that fairness and equity was no longer applicable in the SAS (Canada Revenue Agency (CRA), 2009). However, after several years of operation, taxpayers are well aware of their responsibilities and these issues (fairness and lack of knowledge) have been gradually resolved (CRA, 2009).
Over an extended period, research focusing on complexity of a tax system in SAS has reported that there is an association between complexity and compliance, but the extent of the association remains in part uncertain. Some research has found positive association between complexity and non-compliance, whether intentional or unintentional ( McKerchar, 2002;Ritsema, Thomas and Ferrier, 2003;Blanthorne and Kaplan, 2008) while others have found that the impact of complexity on compliance varied with the characteristics of individual taxpayers; such as income level, education level, perceptions of fairness and equity and the opportunity to evade (Slemrod, 1989). In contrast, Clotfelter (1983) evidenced that when the level of complexity increased (for non-business taxpayers in the case of this study it significantly increased non-compliance among taxpayers. The reason behind this finding was because business taxpayers were more likely to seek advice from tax practitioners as complexity rose; hence the issue of complexity appeared to be significant to them.

CONCLUSIONS
The study concluded that level of tax compliance was affected positively by tax burden and negatively equity whereas complexity of the tax system, fairness, cost of compliance, and policy actions did not significantly affect level of tax compliance. The study recommended reduction of equity and increase of tax burden. Further study was required to assess the effectiveness of complexity, fairness, cost of complying, and policy actions in determining level of tax compliance.