1. Introduction

Finance is an essential requirement for housing development and is seen as the lifeline in real estate development (Tibaijuka, 2002, Ojo and Ighalo, 2008, Onyike, 2009; Kuma, 2015). Tibaijuka (2002) opined that the availability of adequate housing finance is the cornerstone of any valid and sustainable shelter project, while Akinmoladun and Oluwoye (2007) noted that the poor quality and quantity of housing is mainly due to weak and ineffective access to finance. For Onyike (2007) the significance of finance underscores the wish and efforts of every civilized government to see that most of her citizens, if not all have access to housing finance and are adequately housed.

However, access to housing finance is increasingly becoming challenging for prospective homeowners particularly in developing economies. In most cases where homeownership top household's scale of preference, they often forgo expenditures on other domestic consumptions to meet their desired goal (Kuma, 2015). In Nigeria presently, 60% of the new houses developed each year are financed through personal income and savings and sometimes with the assistance from local cooperative societies (National Bureau of Statistics NBS, 2012; Onyike, 2007; Kuma, 2015). Furthermore, a report from the Living Standard Measurement Survey (LSMS) by NBS in 2012 revealed that only 38% of households' have formal bank accounts in South East Nigeria which equal the percentage of the overall national average and only 4% use formal credit in the development of their homes (Kuma, 2015). Perhaps that is why some literature evidence in some case studies indicates that stringent lending policies of formal institutions have influenced borrowers’ choice of finance and as well shifted their preferences to other informal sources of housing finance (Ojo, 2005; Ndibe and Kuma, 2010). This could be the reason many scholars (Onyike, 2007; Onyike, 2009ab; Akinwunmi, 2009; Ojo, 2005, 2004; Atali, 2010) posited that conventional public finance system has failed to offer credit solutions to housing needs of the majority of Nigeria.

Mortgage finance usually has several requirements in which the prospective borrower must satisfy to qualify for the loan/mortgage. The various applicants often perceived these requirements in different perspective given their experience, background and the requirement. For instance, an applicant for loan who is required to have a landed property as collaterals may be reluctant if not discouraged from securing the loan. The repayment scheme and criteria is another factor that may affect the perception of an applicant. It is certain that not all the prospective applicants will perceive the requirement the same. While one may feel collateral requirement as a challenge but to another it could be repayment plan or type of loan provided. Thus, homeowners’ perception of these challenges or factors could account for low patronage of mortgaged finance or banking loan in the study area. As already noted there is a high rate of homeownership through personal savings and other informal sources of finance. Given this, one tend to ask, what are the challenges to housing finance in Nigeria and if there is any difference between perceptions of people who have borrowed and people who have not in the study area. To help answer the above poser, the following hypothesis was formulated. Ho: There is no significant difference between the perception of people who borrowed and those who did not.

2. Review of Related Literature

Literature has identified several factors affecting homeowners’ perception access to housing finance. Ojo (2005), Onuoha (2011) among others have identified the following factors:

2.1. Appropriate Collaterals

(Ojo 2005; Ojo and Ighalo, 2008; Kuma, 2015) found evidence suggesting that collateral such as titled land, government bonds, private shares, properties in urban or rural properties, and fixed deposit are collaterals often considered by financial institutions before granting loans to homeowners. For example, in Nigeria, issues related to land title remain a significant eligibility criterion applied by housing finance institutions for granting loans to prospective individuals who want to develop their properties. Perhaps, this is why Onyike (2009a) posited that the availability of collaterals and other vital inputs determine access to loan affordability. Thus, the decision of a household to finance housing development is dependent on the availability of collateral. However, when trying to borrow loan for housing development, institutional sources of housing finance are often inaccessible to most households especially the low-income earners due to collaterals such as titled land, a condition that has delayed the construction and development of housing.

According to Ojo (2005) issues related to land title in Nigeria remain a significant barrier to housing finance. In many cases, there are often legal impediments to the ability of a property owner to pledge residential property as collateral (Diamon and Lea, 1992; Onuoha, 2011; Kuma, 2015). Again, the process of procuring a registered title to interests in land under the land use Act of 1978 in Nigeria is challenging, tasking, highly expensive and time-consuming. This has not only frustrated the take-off and execution of many viable housing projects but has denied many prospective developers access to mortgage loan due to lack of title collaterals. Ojo (2005) noted that this is very worrisome considering the main thrust of the Act which is to facilitate availability and accessibility of urban and rural land for development in Nigeria.

2.2. Access and Affordability Criteria

Prospective homeowners feel reluctant to borrow loan from financial and mortgage institutions due to stringent conditions on access to loan and affordability. The decision to access loan for housing development in Nigeria and in particular Owerri is dependent on the income of the borrower, income security criteria viewed in terms of stability of employment and business, fixed annuity, method of repayment such as pensions and allowances, equity contribution of the borrower, cost of housing to be developed by individual (Ojo, 2005; Onyike, 2009a; Onyike, 2007; Onuoha, 2011). When one’s income is not even enough to meet one’s immediate needs such as food, financing housing under stringent lenders requirements will not be included in the budget. Onuoha (2011) observed that affordability criteria of financial institutions in Owerri are much hinged on the income profile of the applicants while accessibility is tied to the formal rules governing households' ability to obtain housing loan (Bramley1993; Onyike, 2007). For example, studies show that financial institutions require loan applicants to make a down payment as a condition for loan approval. Also, investigations revealed that low-income earners rarely apply for loans for housing development in Owerri due to the high emphasis on income as affordability criteria (Onyike, 2007; Ahmad, 2009; Onuoha, 2011).

2.3. Repayment Scheme and Criteria

Repayment criteria of a financial institution may well determine the propensity of prospective homeowners to apply for housing finance. As such, the repayment scheme has been observed to be one of the factors underlying a potential homeowner's perceived probability of securing a loan from banks. Applicants who do not agree for a stipulated amortization period for loan repayment are likely to have their application rejected. Ojo (2005) and Onuoha (2011) also found that the probability of homeowners securing housing finance is considerably higher if the repayment is to be made in a short or medium term. Besides, the preparation schedule for loan repayment could have an input in the income of the borrower, interest rate, and loan amount (Ojo and Ighalo, 2008).

In Nigeria, amortization and repayment periods vary with the type and sources of housing finance. Studies have found that repayment criteria of financial institutions are one of the major impediments to housing finance in Nigeria. For example, Onuoha (2011) observed that the payback period granted by banks is too short when one considers that housing development takes longer time to complete. While Ojo (2005) posits that long term loan agreement affords better opportunity for potential homeowners to complete their houses.

2.4. Formal Rules of Access

Studies have shown that experience of those who have attempted to borrow from financial institutions had their loan application rejected or suspended because of what Ojo (2005) and Onuoha (2011) described as formal rules of access. Some financial institutions give additional conditions for loan approval. These conditions are to be strictly observed by prospective loan applicants. These conditions include the age of account with a lending institution, approved building plan, current tax clearance certificate, and evidence of insurance of the property to be financed and bank service charges (Ojo, 2007; Onuoha, 2011). According to Ojo (2005), genuine applicants who wish to develop a house of their own are often denied access to loan due to the short period they have operated an account with banks. Most financial institutions believe that the age of account kept with them builds confidence and ensures quick repayment of the loan. On building plan and presentation of tax clearance, potential borrowers are expected to deposit the architectural plan of proposed building and present evidence of tax clearance to the bank. In some cases, banks require three to five years of tax clearance as a prerequisite for granting loan (Ojo, 2005).

2.5. Type of Loan Provided

Another commonly attractive motivation that determines the propensity of homeowners to initiate housing loan application is the type and nature of loan provided by the financial institutions. Most financial institutions provide short term, medium and long term finance for real estate development (Nwanekezie, 1996; Kuye, 2000). However, every potential homeowner intends to leverage on these categories of real estate finance options to achieve his objectives (Kuye, 2000).

2.6. Type of Project Financed

Subjective evidence suggests that borrowers consider the type of project financed by financial institutions before seeking for a housing loan. Housing construction and development is complex and involves many skills and resources including huge financial commitments. As such, financial institutions hardly provide loans for all the categories of housing development. There is evidence that prospective borrowers may want to borrow a loan for different purposes. For example, a borrower may want a loan for constructing a house or for outright purchase, repairs/renovations and improvement purposes (Kuye, 2000). Nwanekezie (1996) observe that the probability of borrowers securing loans depends to a large extent on the type of project the financial institution finances. It is for this reason that (Ojo, 2005) posit that a borrower is attracted to housing loan if such an arrangement helps him to achieve his purpose.

2.7. Location Factors

It has been said that one most important factor that influences access to housing finance is site and location-related factors. For example, site and location of the property such as neighborhood location, accessibility to popular places of employment, adequate availability of utilities and transportation as well as compliance with applicable federal, state or local land use law and zoning regulations improve the chances of prospective borrowers to obtain housing loan (Nwanekezie, 1996). According to the author, location determines accessibility while accessibility determines demand and consequently profitability. The implication is that financial institutions may be reluctant to finance any project; its location would not add any economic value to the property owner. Perhaps, this is why Ojo (2004) observed that properties located in urban and rural areas are rated differently by financial institutions. For example, banks prefer properties located in urban areas than the rural areas (Ojo, 2005). This is because pieces of property may have different values just because of the difference in location. Moreover, financial institutions perceive that there is general apathy among insurance firms to provide borrowers with mortgage indemnity guarantee insurance premium for properties located in rural areas (Ojo, 2004; 2005). However, it has been argued that when the provision of housing loan is greatly and strictly hinged on location attributes, potential borrowers are often discouraged from accessing the loan (Nwanekezie, 1996). Specifically, those whose properties are located in rural areas or are not in good locations may have their loan application rejected (Ojo, 2004; 2005).

3. Research Methods

The sample frame was drawn from a population of homeowners who borrowed, attempted/or did not succeed in borrowing loan from banks for housing finance. Available data from the regional offices of the banks as of the time of the survey shows an estimated population of 1,500 documented loan applications. A total of 500 sets of questionnaires were distributed among the classes of respondents. For those who borrowed, 180 questionnaires were distributed, out of this number, 150 were returned representing 33.4% of the distributed questionnaires. For those who did not borrow, 320 questionnaires were distributed while 300 representing 66.6% were returned. Questionnaires were administered face-to-face to the participants. Several visits were made, and reminders sent including phones calls to the respondents.

The questionnaire was limited to two parts. Part 1 contained general demographic questions of the participants designed to know their background information. While part 2 comprised a set of questions that were intended to shed light on participants’ awareness of housing finance and on the perception of factors that could influence access to housing finance. The distributed questionnaires tapped into the perception of the homeowners (those who borrowed and those who did not) on the seven (7) identified factors in the literature.

Discriminant analysis was used in this study. The justification for the adoption of discriminant analysis is that the study covers two respondents – homeowners who borrowed and homeowners who did not borrow. Thus, the study involves the analysis of the differences in perception of two groups of sample populations. Discriminant analysis as described by Hair et al. (1987) is a multivariate statistical method used to study the differences and make a comparison between two or more defined groups on a set of the variable measured at interval scale. It involves creating a liner mixture or blending of two or more discriminating variables that differentiate best between groups. This could be done by applying a statistical decision rule of maximizing between-group variance relative to the within-group variance (Ramsover, 1985; Hair et al., 1987; Aliagha et al., 2014). The relationship is expressed as the ratio of between-group to within-group variances, and the linear grouping is a derivative of the following equation:

Z = ₁X₁ + W₂X₂+W₃X₃.....+W_nX_n;

Where Z is the discriminant score, W is the discriminant weights (discriminant coefficients), and X is the independent discriminating variables.

4. Results and Discussion

The data gathered in filed survey were analysed using the discriminant function and the results are presented in tables 4.1 and 4.2.

The Table (4.1) below shows the group means scores and tests of equality of group mean statistics used to identify variables and assess the perception of the two respondents (those who borrowed and those that did not) on factors and variables that affect housing finance. Thirty-four (34) variables were used to determine their perceptions.

Table 4.1. Group Mean Differences and Test of Equality of Group Mean of Factors Influencing Access to Housing Finance in Owerri

Table 4.1 shows that all the six variables measuring access and affordability criteria (AAC) exhibited strong discriminant power, which suggest there were significant group mean differences in perception between the two respondent (those who borrowed and those who did not borrow) on the variables (AAC1 λ 0.986, F = 13.865, p < 0.05; AAC2 λ 0.898, F = 11.108, p < 0.05; AAC3 λ 0.976, F = 11.822, p < 0.05; AAC4 λ 0.998, F = 12.571, p < 0.05; AAC5 λ 0.965, F = 14.431, p < 0.05; AAC6 λ 0.977, F = 16.068, p < 0.05). Though, the two respondents registered very high total mean on the six variables (>4) which overall are considered significantly high. It infers that the variables have strong effects on borrowers’ capacity and access to housing finance. However, the group means values and difference of the respondents were too high to make significant differences. Thus, it could be said that access and affordability criteria variables constitute greater factors that influence housing finance particularly those who did not borrow.

On appropriate collaterals, a look at Table 4.1 indicates that the total mean scores for the six variables measuring the factor range from 4.5 to 3.5 which overall could be regarded as high. This implies that the variables have a strong influence in determining access to housing finance. But a glance at the group mean shows that five variables exhibited strong discriminant power, which suggests there were significant group mean differences in perception between the two respondents on the variables. Furthermore, a look at the Table indicates that those who did not borrow recorded higher group mean value on variable AC5 (4.38) than those who borrowed (3.68), the mean difference was too small to make a significant difference. It suggests that rural properties do not constitute major determinant of access to housing finance and largely both respondents do not see rural properties as key collateral to housing finance.

For variables measuring repayment scheme and criteria (RSC) all the variables showed a very high total mean scores (all >4), indicating that both those who borrowed for housing finance and those who did not succeed in their application or had their loan request rejected recognize repayment scheme and criteria as strong factor that influence access to housing finance. However, this perception is not balanced as those who did not borrow have a stronger view on the influence of the repayment scheme and criteria in determining access to housing finance. Table 4.1 above shows clear evidence that all the four variables displayed strong discriminant power which revealed there were significant group mean differences in perception between those who borrowed and those who did not on the variables. Those who did not succeed in borrowing or had their application rejected recorded higher group mean values (4.93, 4.33, 4.85 and 3.98) than that those who borrowed (3.88, 3.58, 4.30 and 3.34), the mean differences were large enough to make significant group difference. Again, the fact that the repayment scheme and criteria are perceived by those who did not borrow as having a strong influence on access to housing finance requirements is not surprising. This is because investigations suggest that an inability to meet the lending requirement of this nature accounts for 89% of the reasons for loan rejection (Ojo, 2007; Onuoha et al., 2014). Thus, applicants who did succeed in borrowing may be more aware of this having known the practical implications of not fulfilling the conditions for loan repayment on a record time.

Table 4.1 also indicate that six variables determining formal rules of access (FRA) displayed strong discriminant power, which suggest there were significant group mean differences in perception between the two respondents on the variables. The total mean scores of the two respondents are all very high (all >4), indicating that both respondents believe that formal rules of financial institutions affect access to housing finance. Despite this perception, the group means values as shown in the Table 4.1 revealed that those who did not borrow have the stronger perception that formal rules of access measures are more harsh and stringent than those who had their application approved. The mean differences are significant enough to support these differences in perception.

The total mean values for the three variables eliciting measurement for the type of loan provided (TLP) as factors influencing housing finance are all >4 which overall are considered as significantly high. This suggests the variables have a strong influence in determining access to housing finance. It also shows that both respondents agree that this factor determines their motivation to apply for housing finance. However, this perception is not balanced. For example, evidences from the group mean scores show that the three variables exhibited strong discriminant power, which suggest significant group mean differences in perception between the two respondents on the variables. Thus, the type of loan provided constitute a greater obstacle in accessing housing finance by those who did not succeed or had their loan application rejected or suspended. The explanation for this may not be far-fetched. Some financial institutions are often in a hurry to recover their money within a short period. Thus, they provide a short or medium term for their customers. Findings from the study show that borrowers are averse to this arrangement.

Two out of four variables measuring type/purpose of project financed (TPPF) demonstrated strong discriminant power, which suggest there were significant group mean differences in perception between those who borrowed and those who did not borrow on the two variables. However, the total mean scores for the four variables were all > 4 which is considered as high. This suggests that the variables create strong effects for both those who borrowed and those who did not borrow in accessing housing finance.

On the five variables measuring location factors/attributes (LFA), the two group of respondents demonstrated significant total mean scores (all >4), indicating that both respondents regard location attribute as harsh and crucial factors considered by financial institutions before granting housing finance to applicants. However, this perception is considered more stringent by those who did not borrow or those who had their loan application rejected. The five variables displayed strong discriminant power which suggest there were significant group mean differences in perception between the two groups on the variables.

4.1. Predicting Discriminant Function for Propensity to Access Housing Finance and Testing of the Formulated Hypothesis

In order to determine the most predictive variables that best differentiate the two groups and test the formulated hypothesis (those who borrowed and those who did not borrow) the “stepwise method of entering/remove” for deriving discriminant functions was used (Huberty and Barton, 1989; Aliagaha et al., 2014; Onuoha, 2017). Thus, the 34 variables in Table 4.1 above were subjected to the stepwise method. Table 4.2 below shows that at 30 iterations at 0.05 significant level were significant.

Table 4.2. Predictive Model of Access to Housing Finance Variables entered/removed a,b,c,d

The above table confirms the significance of the discriminant function and identified that access to housing finance requirements and variables have the greatest impact and correlation with the discriminant function. The Table reveals a canonical correlation (CCr) of 0.658, which implies that the function explained 65% (CCr²) of variance in the group differences. However, when examining the function’s Wilks’ lambda (Λ), the function is considered significant (Λ = 0:588, χ² (df = 5) = 264.605, p < 0.01). Thus, this study substantively concludes that there is a significant discriminant function that separates the two respondents based on factors that influence access to housing finance.

5. Conclusions

The study concludes that there are significant differences in perception between those who borrowed loan for housing finance and those who did not on the identified variables. It revealed that both those who borrowed and those who did not borrow perceived lenders requirements for housing finance as harsh based on their total mean score and values alone. It is also concluded that measures of access and affordability criteria, possession of registered title collateral (title deed) is perceived as the most difficult of lenders requirements for financing housing from those who have borrow. Other lenders requirements perceived by both those who borrowed and those who did not as stringent and restrictive to the growth of housing in the area are repayment scheme and criteria which include amortization period, moral factor e.g. character of the borrower borrower’s willingness and capacity to pay back (the borrower’s ability to repay the loans is one of the main concerns of the lender's decision), and previous repayment performance of the borrower etc.

The study further concluded that for financial institutions, affordability criterion which some borrowers particularly the individual borrowers who fail to meet the conditions is the income security criteria viewed in terms of stability of employment and business. It is conclusion of this study that lenders perceive that there is general apathy among insurance firms to provide borrowers with mortgage indemnity guarantee insurance premium (MIG).