A Tale of Two Products Part Four: The Performance

INTRODUCTION

In this tale of two products we have looked at the power of housing paradigms on housing microfinance product development and two different product and partnership designs in  Uganda and Kenya. The Uganda product was a flexible housing loan that was delivered by the MFI with institutional support from the Housing NGO, but without branding the NGO to the MFI’s clients or integrating the NGO into the loan process. In Kenya, the product included specific house designs and significantly more construction technical assistance. The Kenyan MFIs product was co-branded with the NGO and staff from the Kenyan Housing NGO  participated in the MFI’s housing microfinance loan process. Having looked at paradigms, products and partnerships, we now move on to performance.

SLOW STARTS

The partnerships between the Housing NGO and MFIs in Uganda and Kenya has very similar start-ups. All of the partners attended the same housing microfinance market research workshop and used similar tools and techniques in their market research and product development. Memoranda of Understanding were signed within one month of each other and the amount of loan capital provided by the Housing NGO was the same. The actual transfer of capital to the Ugandan and Kenyan MFI partners took place at approximately the same time. Although Kenya and Uganda have slightly different operating environments, as East African nations they are relatively similar  when compared to each other, as opposed to Latin American or Asia countries, or even South Africa. The product development process, time lines, capital committed and even the missions of the institutions were essentially the same. The only salient differences were some product features and the partnership design.

Getting from the transfer of capital to actual loan disbursements started quite slowly with each partnership. The MFI in Uganda (designated “UMFI” for this tale of two products) was going through some internal transitions that slowed down the implementation of its flexible housing loan. Implementation of the construction loan at “KMFI” (not actual name) was also delayed in part due to post-election violence. The Housing NGO offered additional institutional technical support to the MFIs to get the products up and running in each country and the MFIs finally started disbursing housing microfinance loans at about the same time..

HIGHLY VARIED PERFORMANCE

Once loans started being disbursed and minor process issues were addressed, performance between the two partnerships became markedly different. UMFI began experiencing high demand for its flexible housing loan and used up the initial loan capital from the Housing NGO. The partnership was refinanced with additional capital, which was quickly put to use. Loans disbursed are approaching 500, but have been controlled due to limitations on available capital.

The KMFI product continued to struggle, despite KMFI being a significantly larger MFI than UMFI. Disbursements never picked up and the partnership was terminated before the loan capital had been fully utilized. Total disbursements of the construction loan during the partnership between KFMI and the Housing NGO were less than 50.

Neither product truly “took off.” The KMFI product performance was significantly poorer than UMFI’s in terms of generating and meeting demand. Even UMFI, however, has failed to take housing microfinance to scale beyond the capital provided by the Housing NGO. It has intentionally limited disbursement to the funds available at a concessionary rate through the partnership rather than seeking additional funds from other sources to scale up and meet the actual demand for the product.

WHAT CAUSED THE DIFFERENCE IN PERFORMANCE?

The partners themselves have various interpretations of their performance. One explanation for the KMFI performance was that the institution focused on and incentivized their core  business loan products rather than the housing microfinance product. Certainly that may be a factor, but I can’t help but believe that having both the Housing NGO and MFI interacting directly with clients may have led to some confusion around the product and between the institutions. We may never really know.

I also believe that  higher  levels of construction technical assistance (particularly using standardized designs) make it more difficult  to generate and sustain effective demand for housing microfinance in Africa. This has been my own experience so far; the greater degree of flexibility of the housing component of the product, the greater the ability to generate effective demand. When comparing higher and lower levels of construction technical assistance in housing microfinance, however, Franck Daphnis has stated that “No empirical evidence currently suggests that one approach is correct and the other is not.” [1] My own hypothesis would be that housing microfinance in Africa with less construction technical assistance will generate more effective demand, a higher rate of loan disbursement and lower cost for the institution. When choosing between a hypothesis from me and Franck Daphnis’ experience, however, Daphnis is the more reliable choice. Even this tale of two products, although true, is largely anecdotal in nature rather than part of a larger empirical study.

CONCLUSION

This brings my tale of two products to a close. The Housing NGO, UMFI, KMFI and their housing microfinance partnerships really exist, although I have not used the actual names of the institutions. The product with minimal construction technical assistance and no direct contact between the housing NGO and the MFI clients did in fact disburse approximately 10 times more loans in the same period than the product with a higher level of CTA and both institutions interacting directly with MFI clients. There were some who predicted this very outcome based on the product and partnership design, even before the first loans were disbursed. Whether that says something definitive about the types of housing microfinance products and their performance is yet to be determined. It is something for more study and I hope that some  empirical studies comparing product types in the African context will be coming in the future as housing microfinance continues to develop on the continent.

[1] Daphnis. F. & Ferguson, B. (eds). (2004). Housing Microfinance: A guide to practice. Bloomfield, CT: Kumarian Press, p. 11.

A Tale of Two Products Part Three: The Partnerships

Two weeks ago I started a tale of two products by looking at two very different housing microfinance product concepts that came out of a product development workshop held in Mbarara, Uganda in 2007. I hypothesized that the difference between the two products could be explained more by the housing paradigms and philosphies held by the designers than by the market research data collected. The workshop was part of an initiative to forge partnerships between a Housing NGO and participating microfinance institutions. Two partnerships were indeed formed, one in Uganda and one in Kenya and the resulting products bore a very close resemblance to the ones developed at Mbarara. Having already looked at the broad differences between the products in a Tale of Two Products Part Two, there are also some notable differences in the partnership structures.

UGANDA

In Uganda, the Housing NGO and the MFI (which I am calling UMFI for this posting) launched a home improvement loan very similar to the flexible housing loan concept that was developed in the training workshop. The Housing NGO developed a market research process, which was implemented by UMFI staff. UMFI and the Housing NGO jointly developed a prototype product and did prototype testing in the field. The Housing NGO’s HMF manager was part of the prototype testing, but was essentially presented to the clients as a HMF consultant working with UMFI and not identified as a Housing NGO staff.

After the HMF product was finalized, the Housing NGO provided wholesale funding to UMFI to launch its housing microfinance portfolio. Promotional materials were also supported in part by the Housing NGO, but the product was branded solely as a UMFI product. The Housing NGO’s name and logo were  not identified with the product to the clients or public. The Housing NGO's HMF manager provided institutional capacity building and worked with UMFI staff in implementing the product, but the Housing NGO did not have any role in the loan assessment, disbursement or collection processes and did not interact with UMFI clients directly except when monitoring progress. Even monitoring was undertaken without branding the Housing NGO’s involvement in the product to the clients.

(click on graphic to enlarge)

KENYA

The Housing NGO’s Kenya process with its MFI partner (which we will call KMFI) was similar to that of their colleagues in Uganda in many aspects. The Housing NGO supported product development, marketing and offered wholesale funds to KMFI. A primary difference in developing the product and partnership was that the product was co-branded, bearing the names and logos of the Housing NGO and KMFI and that the Housing NGO staff were more directly involved in the market research.

Another key difference between the products was the housing components of the loans. While UMFI’s flexible housing loan emphasized client control of the housing component and had little to no construction technical assistance, the Construction Loan offered by KMFI included specific house design options and extensive construction technical support. This resulted in a differing partnership structure based on the premise that KMFI would handle the credit aspects of the loan, but the Kenya Housing NGO would directly provide housing support services to KMFI clients.

The Housing NGO would conduct training on the Construction Loan for KMFI clients at joint meetings with KMFI staff. KMFI clients were trained in the Housing NGO’s housing process and given tips on cost reduction strategies in house construction. The Housing NGO staff would also take part in client assessment visits and help clients decide on an appropriate housing solution. The Housing NGO would then provide on-site monitoring and supervision through the construction process, while KMFI would manage the credit aspects of the loan. The Housing NGO had put a ceiling on maximum loan amount that it would support for each HMF loan and in some cases KMFI added additional capital for clients who were approved for loans higher than what the Housing NGO would support.

(click on graphic to enlarge)

CONCLUSION

The basic product development process and institutional support given by the Housing NGO to its MFI partners in Uganda and Kenya was very similar with the exception of the product being co-branded in Kenya. The nature of the products, however, may have influenced how the Housing NGO would interact with the MFI and its clients once the product was launched. (Or perhaps housing paradigms and an underlying model for partnership influenced the nature of the products.) Direct interaction between the Kenyan Housing NGO staff and KMFI clients as an integral part of the loan process was the major factor differentiating it from the Uganda partnership model during implementation. The question would remain: Given the difference in products and partnership models, would there be a difference in performance?

A Tale of Two Products Part Two: The Products

In A Tale of Two Products Part One, I began to look at the power of housing paradigms on housing microfinance product design. Two groups that attended the same product development workshop in Mbarara, Uganda and had worked together during field research developed two remarkably different housing microfinance product concepts. The workshop was a training exercise that was part of an initiative to forge partnerships between a housing NGO and microfinance institutions. Two rough product concepts that came out of the workshop bore a close resemblance to the final products that were later used in the partnerships; one in Uganda and one in Kenya. This is a continuation of the Tale of Two Products.

As the two product concepts were presented and debated, there was little to no controversy or disagreement over product features such as interest rate, grace period, security or loan insurance. Whether the products should be linked (for existing clients) or stand alone (available to new clients) was also not a major point of contention. The debate (and it was heated, contentious debate) was over what the housing component of the product would look like. Here is a summary of the two housing microfinance products, focusing on the housing component:

FLEXIBLE HOUSING LOAN

The Flexible Housing Loan concept was a home improvement loan. The design concept was to allow the client to personalize the loan to his or her own unique housing situation by proposing the work to be done within the given limits on loan amounts. Loans would be disbursed in cash or cheque and the client would be responsible for overseeing the home improvement project. Clients who used the loan for non-home improvement purposes would have a higher interest rate than clients who used the loan on a home improvement as agreed with the MFI.

The MFI would not provide construction technical assistance to the client. It was assumed that there was sufficient construction expertise locally to undertake typical home improvement projects. This was deduced by the fact that the vast majority of the houses in the area studied during the workshop had good roofs, despite the fact that the households had built on their own using informal construction services that were locally available.

The flexible housing loan product concept aligned closely with the supporter paradigm of housing. It supported incremental building while emphasizing dweller control of the housing process. Existing and informal construction services were essentially seen as part of the solution to be harnessed, rather than part of the problem. Something very similar to the flexible housing loan concept was later implemented by the housing NGO and its MFI partner in Uganda, which we will call UMFI for the purpose of this tale of two products.

THE CONSTRUCTION LOAN

The Construction Loan product concept offered a variety of housing packages to clients. The packages ranged from complete units to designs that could be built in stages with several loans. The objective of the range of house designs was to offer complete housing services for clients with different income levels. The designs had names such as the gold, silver and bronze packages. Loans would be disbursed in-kind or in installments to avoid diversion to non-housing purposes.

The primary emphasis was ensuring the quality of the house product. Full technical services were included, from training of the client in the housing process to supervision and potentially procurement and the actual construction work. The product concept, with its emphasis on a complete solution and the quality of the house aligns with the provider paradigm of housing. A very similar product was launched by the housing NGO in Kenya with its MFI partner there, which we will call KMFI for the purpose of this tale.

THE DEBATES

The two groups at the Mbarara workshop entered into hot debate. What was more important, quality or flexibility? Could the institution providing the loan accept a housing result not up to given quality standards? Did the institution have the capacity to deliver construction services at scale? Did clients have the ability to manage their own construction process? In the end, the debates were not resolved and the workshop facilitator had to close the session.

Several months later, UMFI went on to offer a flexible housing loan in partnership with the housing NGO. At almost the same time,  KMFI offered something similar to the construction loan with the same housing NGO in Kenya. Aside from the clear differences in the product features, how the housing NGO and MFI interacted with each other (and with the MFI clients) was markedly different in Uganda as compared to Kenya, but that is for another posting.

Housing Paradigms and Housing Microfinance

I had been working in the housing sector for 15 years before I came to realize that there were two very different interpretations of what housing is and how to approach it. I originally assumed that there was a housing problem (especially for the poor) and that the natural solution to it was to build low cost houses. Over time I began to realize that effectively building houses for the poor held a lot of inherent challenges and the response was not always what was expected. Sometimes people didn’t really like the low cost houses that had been designed for them. They also often did not like to pay for them. Demand was much lower than we anticipated when offering the opportunity for a “good house” to people who were living in  materially very poor houses. I was experiencing symptoms related to the “provider” approach to housing and a focus on the house itself rather than the house in the context of its dwellers' lives.

In Freedom to Build, John F.C. Turner described two ways to define housing: Housing as a noun and housing as a verb. Housing as a noun refers to the physical structure: The house as a product or commodity. Housing as a verb focuses on the universal activity housing.[1] The two definitions correspond to two different ways of looking at housing. Those who view housing primarily as a noun will focus on physical housing units. Those who view housing as a verb are more likely to look a housing as an on-going process and concentrate on the role of housing within the context of the household's broader livelihood. Turner’s second law of housing (see previous post) states that what is important about housing is what it does and not what it is. This could be seen as a sort of manifesto for those who see housing as a verb. How we look at and define housing ultimately shapes how we approach housing interventions.

Nabeel Hamdi built upon the idea of housing as a noun or verb in his book Housing Without Houses: Participation, flexibility and enablement. Hamdi outlined two paradigms of housing that are often in conflict: Provision and Support. The provider paradigm holds that the solution to housing deficits is to build houses. Providers tend to control the housing process to deliver housing units completed to a certain standard. Rather than controlling the production of units, Supporters look at the management of resources such as land, services and finance to assist dwellers to improve their housing, rather than controlling the production of units. Providers and Supporters differing approaches can be seen this diagram adapted directly from Housing Without Houses[2]:

(click on graphic to enlarge)

Someone (or an institution) can be identified as a predominantly a provider or supporter by the language they use and the interventions they design. Providers place a heavy emphasis on their role in building housing units. They maintain a significant level of pride and ownership in the finished physical product. Their interventions tend to keep control in the institution's hands and focus on the standards of the units. The houses produced are almost always complete units. Even when implementing an incremental building scheme that uses Supporter-like language, the Providers' incrementalism is often controlled and built in complete, stand-alone stages to a given standard . Given the choice between their standards and dweller choice that may compromise those standards, Providers will stick to the standards. Providers tend to see those engaged in informal housing services commonly accessed by the poor as suspect at best, but often as illegitimate actors in the housing environment who are a menace or the source of the problem.

Supporters are often a little more ambiguous about what they actually do, because there are a wide variety of possible support interventions and approaches. They usually do not, however, see themselves as builders and instead leave significantly more control of the housing process in the hands of the dwellers. Incremental building is accepted as a reality and part of a housing process. As such, supporters are comfortable with interventions based on progress that does not necessarily result in a "complete housing solution" to the Provider's standards. Whereas Providers tend to see housing as a problem to which they bring the technical expertise to solve, supporters believe that “most solutions exist in everyday practice, they only need to be recognized and the built on. They exist not as governments and professionals might like and might not be working as effective as we would need them, but they exist nevertheless.”[3] This results in supporters usually embracing informal housing activity rather than doubting its legitimacy.

Because two sides of the divide interpret housing differently, they do not see eye to eye when it comes to their respective interventions. Supporters are likely to question the ability of Provider interventions to sustainably serve low income households at scale. They may also raise concerns whether the Providers' solutions are appropriate and sustainable in the context of the dwellers' livelihood strategies. Providers in turn inevitably will point out that Supporter interventions do not meet their housing quality standards. It is at its roots a philosophical conflict on the nature of housing that can sometimes resemble dialogue and debate between opposing political parties.

I doubt that any person or institution holds a 100% Provider or 100% Supporter position. The paradigms are more like a continuum with provision on one side, support on the other and a range of positions mixing the two in between.

(click on graphic to enlage)

Now… this seems all highly theoretical and esoteric, but the housing paradigm of the implementer has a tremendous and visible effect on the design of housing microfinance products and services. The degree to which an institution focuses on the house that results from a housing microfinance loan is an indicator of its housing paradigm. I have heard staff from an institution that provides housing microfinance berate housing choices made by their clients. They insist that the institution must only support "quality" houses with their loans and  must therefore exercise greater control on loan use. The implication was that it was more important for the resulting house or home improvement to meet institutional criteria for satisfaction than the client’s own criteria for his or her housing process. This is a view from the provider paradigm. I have also heard others deeply question whether construction services attached to a housing microfinance loan were sustainable or even necessary: The view from the supporter side.

HMF Hypothesis One: The degree to which an institution engaged in housing microfinance holds a provider or support paradigm of housing will be evident in level of construction technical support offered as part of the product.

One key factor differentiating housing microfinance products is the level of construction technical assistance offered. Some products bear the name housing microfinance or home improvement loan, but in practice are little more than a consumption loan with the word housing tacked on. Other housing microfinance products exert significant control over the loan use and the clients' housing process. Putting the above hypothesis into the proposed housing paradigm continuum, it might look something like this:

(click on graphic to enlarge)

HMF Hypothesis Two: More CTA (provider approach) will result in a higher likelihood of achieving quality standards in the house resulting from a housing microfinance product, but the associated costs and complications in the delivery process will make it harder to achieve sustainability and scale while serving low income households (the double bottom line). Less CTA will make it easier to reach sustainability and scale, but will give less of a guarantee on the quality of the house or home improvement for which the HMF product was used.

(click on graphic to enlarge)

I see housing microfinance as a support intervention. In (what I believe to be) its purest form, it is a finance intervention that provides capital to households so that they can address a finance gap in their housing process. It corresponds to housing as a verb when it leaves control of the housing decisions to the dwellers. When housing microfinance is approached as a support intervention, I believe it has the highest likelihood of reaching scale and sustainability while serving low income households. As a housing microfinance product leans more towards the provider paradigm, I would expect 1) the delivery process to become more  complicated and stress capacity, 2) demand to decrease as dwellers lose control of their housing process and either 3) subsidization to make it affordable to households with low incomes (at the expense of sustainability) or 4) an "up market" drift to wealthier clients to recover high delivery costs (at the expense of social performance).

It is not only the housing sector that may tend to lean towards the provider paradigm. Even some MFIs envision housing microfinance that more resembles a beautiful house with a conventional mortgage , than affordable housing finance for low income households  engaged in an incremental building process. I believe effective housing microfinance for households around the poverty line will fit into the housing process in which they are already engaged, but that is another topic.

I welcome any feedback through comments on these concepts and working hypotheses, particularly opposing viewpoints. I would also like to recognize once more the work of Nabeel Hamdi in describing the housing paradigms and John Turner's concept of housing as a verb. Although liberally interpreted and applied, I am indebted to their work for its influence on my approach to housing and housing microfinance.

[1] Turner, J. & Fichter, R. eds. (1972). Freedom to Build, New York: MacMillan. 151.[2] Hamdi, N. (1995). Housing Without Houses: Participation, flexibility, enablement. London: IT Publications LTD. 27

[3] Hamdi (1995). 36