Effects of Rent-Controlled Public Housing on the Supply of Private RentedHousing and Household Mobility in Malawi: A Case Study of Malawi Housing Corporation

Despite global policy shift towards liberalization of the housing market, different forms of rent control have still remained in a number of countries. Most of these controls are currently uphold by the view that they are different from the old forms that involved complete freezing of the nominal rents. Generally, this view has been sustained by lack of in-depth studies on the remaining forms of rent control. This paper fills this gap by examining the effects of rent-controlled public housing sector on the supply of private rent housing and household mobility in Malawi. The findings from Malawi demonstrate that the effects of rent control are varied and depend on the nature of residential market as well as the political environment under which the rent controls are implemented. Under dictatorial regime, rent controlsin Malawi produced negative effects on both the supply of private rent housing and household mobility. This is consistent with existing studies from many other countries. However, after adoption of liberal democracy and economic policies, rent controlhas produces varied outcomes. While its effect on constraining household mobility remains undisputable, rent control in the public housing sector has insignificant effect on the supply of private rented housing. These differences may be explained by the fact that during dictatorial regime private housing sector in the country was also de facto under rent control.


Introduction
This paper undertakes a critical review of the effects of rent-controlled public housing sector on the supply of private-rented housing and household mobility in Malawi. It has been inspired by the fact rent controlis back on the policy debate in both developing and developed countries almost three decades after its abolitionin a number of developed countries [5; 2]. There is no doubt that despite unanimous and strong criticism against rent controls from mainstream economists [36; 7; 18; 4], divergent views on the subject do not only exist, but are quickly emerging into a majorpolicy debate. For instance, Arnott [2] explicitly called for a revision of stand over rent controls, arguing negative perception of rent controls is misplaced because of confusion between first and second generation rent controls. For instance, in UKthe Labour party was in 2015 considering some form of private rent cap once elected 1 . This turn of events equals what Malpezzi and Ball [23: xi] pointed out that little is known about "different regimes that diverge from text book model" of rent control. The Malawian situation is an example of such divergent rent control regime that merit attention. Despite adopting liberal economic policies in the middle 1990s, Malawi's public housing has remained under rent control andrentals have remained very lower than the market rent. Malawi Housing Corporation (MHC), the only institution providing affordable housing, has for several times been prevented by the state and the courts from raising rentals to market rent. For instance, in 2017 the state allowed MHC to raise rental by a small percentage. All this is happening at a time Malawi is urgent need of housing and where broadly the market is entrusted to be the main Sane Pashane Zuka: Effects of Rent-Controlled Public Housing on the Supply of Private Rented Housing and Household Mobility in Malawi: A Case Study of Malawi Housing Corporation instrument of achieving that goal. It is thus the aim of this article to examine the effect of rent control on the provision of private-rented housing and household mobility. Basing on the supply and demand force analysis and experiences from a number of countries that had implemented such a policy such as USA, Britain, Ghana, and India, the article concludes that the effect of rent control on supply of private-rented houses varies with the broader macroeconomic framework.

Definition and Regimes of Rent Control
Rent control can generally be defined as a restriction on landlord freedom to increase rent paid by the tenant beyond a determined level [30]. It is types of state intervention into housing pricing with a view to typically protecttenants from opportunistic landlords during times of rising rents [2]. There are various and at times confusing regimes of rent controls. One of the most encompassing classifications of rent control regimes is provided by Arnott [2] namely: first generation rent control and second generation rent controls. The former were historically implemented during first and second world wars while the later were implemented after this period. The major difference between these forms of rent controls is that the first generation rent controls were an absolute freeze on nominal rent increase while second generation allows restricted rent increase [4; 18]. Malpezzi and Ball [23] also broadly divide rent controls into control of rents and control of rent increase. Control of rent is concerned with fixing fair rent for the housing unit and may or may not allow future rent increase. On the other hand, control of rent increase involves regulation of future rents. As with the other classification, the major difference between the two is that the lateris an absolute freeze on nominal rent increase while the former involves determining "fair" current rent.

Rent Control, Housing Supply and Household Mobility: A Theoretical Discussion
The key theoretical question in rent controls is its ability to achieve set out objectives in a housing market. The dominant view is that rent control distorts the housing market and consequently results in a number of unintended outcomes such as shortage of housing supply, restrained tenants' mobility, deteriorating housing quality, increases in bribery and increased search costs [8]. This is based on the idea that the practice does not promote private investment into the housing sector and increases the tenant's opportunity costs of engaging in bribery and household decision to change housing [8]. The relationship of rent control and housing supply is specifically analysed in terms of two aspects namely: rent control as a price control and rent control as expenditure control. The former aspect viewsrent control as a tax on the investment returns of the investors which reduces the rent landlords charge per unit of housing services [23].  From figure 1, rent control results in price movement from Pe to Pmax and consequently resulting into excess demand of Q2-Q1 that is not served. Consequently, reduced profit returns discourage landlords from investing in private renting housing which result in less housing being supplied than demanded, hence creating private renting housing shortage. As in all other price controls, the net gain to both tenants and landlords is reduced due to the deadweight loss that is not recovered by neither of them [18]. This theoretical explanation is in line with empirical evidence from a number of studies that have implemented rent control [29; 6; 14; 18]. Malpezzi and Ball [23] specifically found that "countries with no or weak rent controls invest about 6% of the GDP on housing, on average, while countries with strong controls invest 3-4 percent on average". The second aspect sees rent control as expenditure control in which less profits force landlord to reduce the quantity of housing services. According to this understanding, the direct result of rent controlis low expenditure in housing maintenance [29]. Thus, both lower rents and maintenance expenditure demotivate landlords to carry out housing maintenances [18; 4].
There is increasing consensus that the effects of rent control on the uncontrolled sector arevaried and may not be easy to define. For instance, while Fallis and Smith [14] found that rent control can hold down prices in the uncontrolled sector; other studies have strongly indicated that rent control do result into increased housing prices in the uncontrolled sector. Fallis and Smith conclusions were based on their studies in Los Angeles and California. With regard to the second view, Malpezzi [24] found that in Egypt, rent control resulted in the increase of rental levels in the uncontrolled sector. Malpezzi and Ball [23: 54] specifically highlighted that "rents in the uncontrolled sector were much greater than predicted by the model. The average predicted rentto income for this group was 0.16; the actual observed was 0.53". The general theoreticalfoundations explaining the effect of the controlled sector on the uncontrolled one is that prices in the uncontrolled sector are driven up by excess housing demand spilling over from the controlled sector [4; 18; 2; 14]. However, Malpezzi [24] argues that in the long run prices of the two sectors are likely to converge.
With regard to household mobility, both theoretical foundations and empirical evidences on rent control strongly indicate that sitting tenants will be less mobile as the opportunity cost of moving will usually be deceivingly high for them to bear. Sitting tenants will specifically consider the difficulties of finding a similar or better rent-controlled housing unit or forgo the subsidized unit they are now occupying for a higher rent [18]. Generally, this considerationis the source of a number of inefficiencies as housing units are not efficiently allocated and sitting tenants fail to take up promising new opportunities elsewhere because they want to maintain the currenthousing unit [3]. For instance, Block and Olsen [8] found that fear of losing rent-controlled tenancies made Viennese workers in Austria fail to move from their residential area to an employment abundant neighbouring manufacturing area.

International Experience with Rent Controls
Rent control has a long history at international level. In the developed countries of USA, Britain, Sweden and Israel, the introduction of rent control is associated with the two world wars. For instance, in Britaincomplete rent freeze was implemented between 1915 and 1920 and between 1939 and 1954 with the passing of the War Restrictions Act 1915 and Rent and Mortgage Interest Restriction Act 1939 respectively [2; 6]. In USA rent freeze on nominal rent was implemented between 1939 and early 1950s [3]. In these countries, rent controls were a temporary strategy of providing affordable housing to poor households and prevent landlords from profiteering from the wartime housing shortages. This situation was caused by rising inflation and a complete halt of housing construction due to war commitments. Most of these countries adopted first generation rent controls but have now either deregulated private housing such as in Britain or have adopted second generation rent control such as some states in USA [18].
Unlike in developed countries, rent controls in developing countries were introduced to manage the housing challenges that came with either Second World War or urbanization. For instance, rent controls were introduced in Ghana in 1942, Egypt in 1944, India in 1947, Philippines during early years of World War II [25; 15]. Historically, all these countries had very strong links with European countries that were at war such as Britain. However, in some countries such as Malawi and Zambia, rent controls were introduced in the middle of the 1960s as a way of subsidizing the cost of housing for the urban civil service [35; 36]. In Malawi in particular rent control isapplied only to the public housing sector i.e. Malawi Housing Corporation [19; 34]. To this end, varied effects and outcomes of rent control are expected.

Residential Housing Market, Malawi Housing Corporation and Rent Control in Malawi
Residential housing market in Malawi is heavily dominated by informal individual household investors, a few medium scale investors and Malawi Housing Corporation as a public housing sector [35]. This market structure is largely a product of the extent to which the state actively participated in the provision and regulation of residential housing. Recognising the need for housing, the government created Malawi Housing Corporation (MHC) at independence to act as a state agent in delivering rent-controlled public housing to civil servants and low income groups [28].
MHCwas mandated to constructhouses for rent and sale, service and saleplots to private developers and manage existing residential portfolio. According to the Malawi Housing Act, 1964) [17], the institution is supposed to operate as a non-profit making organization providing affordable housing. It is, however, allowed to increase rent in line with the increasing cost of housing construction and upon government approval. However, since independence government has reluctantly allowed MHC to raise rentals and in cases where it has been allowed, the increase has been below the cost of servicing the houses [34; 35].
Since independence, the private residential housing sector has not been under formal rent control. It should, however, be noted that although the state did not explicitly control private rented housing, active state involvement in the economy coupled with active state control of commodity prices rendered private sector housing de facto under rent control [9]. This is so as during the period 1964-1990s, the state informally but effectively intervened in situations where private sector prices differed widely with its preferred prices. Thus, the widely held expectation was that rentals in the private sector would not differ so much with those in the public housing sector [33]. Private housing in Malawi was, therefore, de facto under rent control from independence to middle 1990s when the country adopted multi-party democracy and liberalized economic policies.
Like in many countries with a history of rent control, liberalization of the economy in Malawi did not correspondingly result in rent de-control. MHC remains a parastatal institution offering rent-controlled affordable public housing and despite calls for rental de-control, all government regimes that have ruled the country since the advent of multiparty democracy have dogged the question of reforming the Act to enable MHC charge market rents [19]. Currently, MHC is charging almost half the prevailing market rent. For instance, a three bedroom house offered at a rent of US$100 under MHC is charged US$200 under the private residential market. Generally, government failure to abolish rent control in public housing in Malawi depicts international experiences and dilemmas of abolishing rent control. In a number of countries such as USA, Britain, Ghana, India and Egypt, some form of rent control have remained long-time after policy change in favour of market based pricing [18; 23]. In Malawi, government reluctance to provide legal framework for the implementation of market rent in the public housing sector has resulted into all attempts by MHC to effect charge of market rent facing public resistance and as in line with the existing law effectively been quashed by the courts [16]. For instance, the courts concluded in one case of MHC versus its Tenants that it was "clear from the wording of the Act that in carrying out its operations, the Malawi Housing Corporation's purpose was simply to break even, not to make a profit" [21: 147].
Thus, the history of rent control for the public sector in Malawi presents two contexts for analyzing the question understudy.
Firstly, rent control during state-led development period when though rent control was applied only to MHC affordable public houses, prevailing protectionist economic ideology made the housing sector de facto under rent control. Secondly, rent control after adoption of liberalized economic policies when rent control is effectively applied to MHC affordable public housing and not the private residential sector.

Methodology
This paper is based on secondary data and residential market survey in the city of Blantyre, Malawi. Blantyre was particularly chosen because it is the commercial city of Malawi and has the highest number of MHC houses. Secondary data was collected from MHC, private real estate companies such as MPICO, Knight Frank and other publications on housing in Malawi. Residential housing survey was conducted in Blantyre and focused on prevailing rentals for the private sector in different locations. The study randomly sampled and interviewed 80 tenants renting private residential housing units in different locations in Blantyre.

Effects of Rent-Controlled Public
Housing on Supply of Private Rent Housing

During State-Led Development, 1964 -1990s
One of the major effects of controlling rents in the public housing sector in Malawi is low supply of private rented housing. According to UN Habitat [34], for a long the supply of private rent housing in Malawi has not matched the ever growing housing demand especially in the face of high urbanization rate of around 6.3%. This trend is against a harsh reality that MHC, a vehicle entrusted with the responsibility of providing rent-controlled housing, has been in operation for over fifty years. For all this period, MHC serves less than 20% of the households living in rented housing in four major towns [19]. According to Manda, Nkhoma and Mitlin [26], MHC has failed to meet the housing demand because government failed toprovide subventions required for its development investment and by 1981 the institution stopped constructing housing units. Kadale Consultants [19] particularly highlights that the rentals that the institution received was not adequate to cover its operations and the institution is only constructing a maximum of 300 housing units. All this is coming against a reality that the country will need about 254, 500 new dwellings by 2020, translating into 21, 000 new houses required to be built each year [34].
Shortage of private rented housing in Malawi can be explained by lack of investors' interest to fill the supply gap created by dwindling housing supply from MHC [34]. In particular, existing rents could not attract private investors as rental income could not cover developers' construction costs and the required return. Thus, apart from state-financed institutions, very few private investors invested in the sector between 1964 up to the adoption of liberalized economic policies in the 1990s [19]. For instance, Malawi Property Investment Company (MPICO), the country's biggest commercial property company, sold off its entire property portfolio in the residential sector in the 1980s because of low rentals and high maintenance costs [19].
It is this shortage of housing stock that has forced urban dwellers over the years to self-build in unplanned settlements. It is reported that about 40% of the urban population currently live in their own house; and the population living in unplanned settlements has increased from 45% in 1977 to 60% in 2009 [34; 20]. It should, however, be stated that for most households owning a house in town is a temporal coping mechanisms as "migration to urban areas rarely results in permanent settlement" [13: 137]. Thus, the effects of rent control in Malawi support the conventional wisdom that widely applied rent controls result in reduced supply of private-rented housing as developers fail to recoup construction costs and achieve the targeted profit returns [7; 8]. These findings are consistent with experiences of a number of countries that have experimented with this policy such as Britain and USA [4]. In Britain, percentage of households living in private rented housing fell from 61% to below 15% [3].

Effects of Rent-Controlled Public Housing on Supply of Private Rent Housing After Liberalization of the Economy -1990s to Present
The adoption of liberalization economic policies in the middle 1990s changed the market dynamics in the residential sector in Malawi. Rental levels in the private housing have bounced back very strongly and housing prices are increasing at a faster pace than expected [34]. Using the cost of construction as a proxy indicator for expected housing prices increase 2 , UN Habitat [34] indicates that during the period 2000 and 2003 the cost of constructing a standard 3-bedroom house rose by 3.18 times while inflation during the same period rose by 1.54 times. Consequently, rental levels in theprivate rented housing sector have more than doubled those in the public housing sector [34]. This situation is worse in Lilongwe, the country's administrative city, where high population growth and increased economic activities is driving rental levels very high. According to Knight Frank [22], rents for a 4 bedroom house in Lilongwe are twice as high as in Blantyre, the country's commercial city -US$1700 to US$3500. While there are many factors explaining the differences between the two centres, it is interesting to note that Lilongwe has less numbers of MHC public housing units than Blantyre [34].
The rate at which housing prices are increasing in Malawi and the number of private investors that have entered the market since early 2000s indicates that supply of private rented housing was artificially held back by rent control in public housing sector and its defacto application in the private rented housing sector. While medium to large scale investors existed the residential housing market between 1964 and early 1980s, the 2000s have registered both the return of old investors and the entry of new investors such as Pacific and Press Trust [19]. Equally important, regional property companies such as Knight Frank and Brollhave established their property management office branches in the country in response to increasing returns in the sector. In 2013, Knight Frank [19] documented residential sector prime yields as low as 4.25%. It can, therefore, be concluded that unlike in the state-led development period, rent-controlled public housing do not currently have significant effect on the supply of private rented housing.
However, the combined effect of shortage of housing units during the period of private sector rent control and high rental after liberalization has resulted in high demand for home ownership. About 60% of the people interviewed indicated that renting is a temporary alternative but their desire is to build and own a house of their own. While home ownership gives home occupiers economic stability, it does not promote the growth of real estate management services. Findings from this study points to stabilization of housing rentals as relative demand for housing units falls as a result of increasing proportion of home ownership.

Effects of Rent-Controlled Public Housing on Household Mobility in Malawi
Just like in many other countries that have implemented rent control, limited household mobility is one of the major outcomes of rent control on public housing in Malawi. Findings from Malawi indicate that during both periods 2 The assumption was based on the idea that house price in the country is mainly affected by cost of construction (before and after liberalization of the private housing market) most tenants stay in public housing units for a long time and constantly fight against their removal and rental increase [16; 21]. This practice is reinforced by the fact that sitting tenants are given first priority to buy out the housing unit they are occupying before it is put on to the market [17]. Thus, both the benefit of paying low rent and expectation of buying a cheaper house motivate sitting tenants to remain in the house for a long time. This study has found that some tenants informally sub lease their housing unit to their relatives or friends in order to maintain the housing unit. However, it is unlikely that the housing demands of most people requiring affordable housing will be realized soon given that MHC isnot constructing enough houses and has 100, 000 households currently on its waiting list [19; 27]. This figure has steadily been increasing from 7000 in 1971 to 32000 in 1982 [19]. Tenants generally prefer MHC houses because they are not only cheaper compared to private rented housing units, but have also flexible rental payment schedules. For instance, MPICO low cost houses failed to sell because those who could afford to buy them preferred low density areas as offered by MHC [19].
The findings from Malawi resonates well with theoretical underpinnings that rent control potentially worsens the welfare of the people it is intended to protect by, inter alia, subjecting them to deteriorating housing quality, increasing bribery and passing over of maintenance costs from landlord to tenants [31; 8]. Jenkins [18: 77] particularly argued that rent control results in misallocation of housing units as "it is not necessarily the highest benefit users who get in -tenants may apply for or remain settled in the apartment that do not well suit their needs simply because the apartment carries low price". Paradoxically, strong opposition to public housing sector rent decontrol by both MHC sitting tenants and government in Malawi seems to indicate that the system is serving its purpose and thus benefiting the low income groups. Experiences from Malawi point to the view that this perception might be flawed as it may not be based on adequate cost-best analysis of the housing services that thesitting tenants receive from MHC [3; 8; 18].
One clear point to illustrate this point is the fact that MHC failure to carry out routine maintenance of its housing portfolio has forced sitting tenants to do almost all of the maintenance on their own [26]. This practice is obviously costly to sitting tenants in an environment where annual inflation on routine house maintenance is rising. For instance SADC [32] indicated that inflation for routine housing maintenance in the country rose from less than 1% in 2011 to more than 5% in 2013 [32; 1].
Equally important, government unwillingness to facilitate decontrol of public housing is characterized by sheer deception. In particular, the motive of government is to placate low income groups into believing that public housing is largely in their interest. On the contrary, public housing is effectively serving the interest of politicians and high income bureaucrats by enabling them buy houses at highly reduced prices. For instance, in 2011 MHC sold houses to top politicians and MHC senior staff at very low prices and made a net loss of about MK100 million (around US$640 thousand in that year) [10]. This finding mirrors a case in USA where a Mayor of New York retained his rent-controlled apartment longtime after he became mayor [12]. Equally important, the housing services of MHC housing units may be costly compared to private housing units if land area and existed of modern facilities is featured into the analysis. These findings, therefore, reaffirms fears of a number of economists that rent control is inefficient in allocating resources to those that are in need [18]. Thus, understanding the full effects of public rentcontrol in Malawi, and other countries with similar political environment, demandsboth econometric and political economy analysis to identify the full range of actors involved in rent-controlled public housing sector as well their interests in and benefit from the system [11]. The findings from Malawi seems to indicate that while rent control in the public housing sector is supported among the low income group, benefits accrue to the elite especially ruling politicians, bureaucrats and MHC senior staff.

Conclusion
This paper was set out to examine the effects of rentcontrolled public housing sector on the supply of private rent housing and household mobility. Findings from Malawi demonstrate that the effects of rent control are varied and depend on the nature of residential market as well as the political environment under which the rent controls are implemented. Thus, it is not easy to define priori the effects of rent control in each country as it largely depends on the existing economic and political landscape.
On the overall, the evidence from Malawi is consistent with the view that widespread rent control results into the reduction of housing supply and constrain household mobility. In Malawi particularly, de facto application of rent control to the private sector during one party era effectively curtailed private investment into residential real estate sector. Consequently, private housing supply was very low leading to increased selfbuild as a coping mechanisms for household shortage. Equally important, household mobility was constrained as siting tenants tried to cap their housing expenses by holding on to the cheap housing units they were occupying.
However, rent control has produced mixed outcomesafter adoption of liberal democracy and liberalized economic policies. Specifically, rent control on public housing has had insignificant effect on the supply of private rented housing. Unlike during dictatorial rule, this period has seen new investors entering the market as well as the return of old investors. On the other hand, household mobility in the rentcontrolled public housing sector has not changed. Sitting tenants are not only holding on to their housing units, but are also fighting any initiative to bring rent closer to market rental levels. The different outcomes may be explained by the fact that during dictatorial regime private housing sector in the country was de facto under rent control. These findings, however, need to be properly contextualized. In particularly, low production of rent controlled public housing units in relation to demand means that the controlled sector is almost non-existent and, therefore, has insignificant effect on the demand-supply dynamics of the private housing sector. Currently, MHC produces less than 300 housing units per year against anestimated demand of 21,000 units. Predictably, these outcomes may be different where there is a performing public housing sector.