Rental Legislation and the Changing Icelandic Rental Market

Rental markets play an important role for a functioning housing market. Households with shorter time horizons, as well as households with little private equity and difficulties acquiring capital can see their housing needs met on the rental market. A functioning rental market does however require some sort of legislation. In this paper we argue that legislation must adapt to changing rental markets. We do so from the specific case of Iceland. The Icelandic rental market has, since the financial crisis, undergone noticeable structural change, evident from a number of perspectives. Based on the presented structural change of the Icelandic rental market, and our presented view on the role of private rental housing, we conclude that there is a case to be made for changes to the Icelandic legislation.


Introduction
Rental markets can cater to different types of individuals and households depending on what type of legislation a country has implemented. However, the type of households who turn to the rental market can change over time regardless of the structural legal framework. If this happens, it may warrant changes in the legal framework. The rental market in Iceland, we will argue, has undergone structural changes since the financial crisis of 2008 (from here on referred to as the financial crisis). More households have become renters, and they stay renters for longer. In addition, the number of professional for-profit rental companies has increased and the majority of their market share is divided between two large companies. With this paper, we aim to present a discussion of the legislation governing the Icelandic rental market. The question of pressing interest is: Is there a case to be made for changes to the legislation in order to adapt to the evolutionary process of the private rental market?
Icelandic housing policy was centered around a home-owning society with minimal state involvement until the 1980s. Thus Icelandic housing policy differed greatly from that of the other Nordic countries, especially Sweden in the post-World War II era (Sveinsson, 2004;Bengtsson et al., 2013). Innovations in the https://doi.org/10.30672/njsr.95233 fisheries sector in the early 20 th century fueled an economic change which caused rapid urbanization, forming a large rental market in the fast-growing towns. By 1920, half of the population in Iceland lived in urban areas, a share which had more than doubled from the beginning of the century. 53% of housing in Iceland were rentals. The rental market was largest in Reykjavik where it is estimated that 63% of all housing were rental apartments. The share of rentals remained high up until World War II. Homeownership in early urban Reykjavik was highly related to class and wealth (Sveinsson, 2004, Statistics Iceland, 1997. However, the share of rentals decreased after World War II and reached 17% in Reykjavik in 1983 and 11% for the whole country in 1990 (Statistics Iceland, 1997). The allied occupation of Iceland brought about strong economic growth during and after the World War II, and resulted in improved standard of living and a sharp rise in per capita income. This can probably partly explain this increasing share of homeownership. However, exceptionally high inflation from  is likely to have played a large role in this development (Sveinsson, 2004) as owning your own house provided protection of your savings (Ministry of Social Affairs, 2004).
The current legislation governing the rental market in Iceland came into effect in January 1995 as a revision of the former legislation from 1979 and has not undergone major revisions since then. The legislation from 1979 was considered in many ways unsuitable for the Icelandic market (Ministry of Social Affairs, 2004), however it provided some tenant security. For the most of the post-war period, following war time rent regulation, the rental market was more or less free and unregulated (Sveinsson, 1992).
Following the new and revised legislation attitudes towards renting started to change and laws on rental assistance also played a role by inducing improvement in the condition of rental properties (Ministry of Social Affairs, 2004). Prior to the financial crisis in 2008, the rental market stayed relatively small and policy changes at the time further stimulated a rise in homeownership. In 2004 the Housing Financing fund, a state-owned mortgage lender, eased its loan regulation and started offering 90% mortgage loans. According to Elíasson and Pétursson (2009) this caused a strong response from the newly privatized domestic commercial banks that actively entered the mortgage market for the first time. Long-term real mortgage rates declined, access to credit increased and housing demand rose substantially, causing a 25% increase in house prices one year after the shock and similar percentage rise in housing investment two years after the shock.
Since the financial crisis the size and structure of the rental market as well as attitudes and preferences towards renting, seem to have undergone change. Government policy has started to cater more to the rental market than before for example by providing increased housing allowances and by incentivizing the construction and purchases of rentals for lower income households by providing endowment capital to non-profit companies. There has also been an increased political pressure and demand for new legislation. This is the focus of the current paper.
The rest of the paper is structured as follows. Section 2 will provide an overview of the role of rental markets. Section 3 will provide the argument for a structural change of the Icelandic rental market. In section 4 we provide a discussion of the Icelandic private rental legislation and how changes may be motivated in line with the structural changes presented in section 3. Section 5 provides a discussion of the results, and concludes.

The role of the private rental market
In order to provide a fruitful discussion of changes in the private rental market and associated changes in legislation governing said market we need some type of understanding of what the role of the housing market is. Such perspectives can be drawn from different research disciplines. In this paper we will mostly focus on the market perspective given by the economic science, although we will briefly discuss non-market based viewpoints stemming from other social sciences.
The most obvious prerequisite for a functioning rental market is the existence of apartments available for rent. For the purpose of this discussion we will assume that this is already in place. If, for some reason, policy makers want the rental market to increase, this can be achieved in a number of different ways. For example by changing land policies or by introducing subsidies to new construction. Such political instruments do however lie outside the scope of the current paper. In addition, the structure of rent legislation will likely differ depending on the structure of the rental market. Kemeny (1995 and2007) (further discussed in e.g. Kemeny, Kersloot and Thalmann (2005)) has developed the concepts of 'integrated' and 'dualistic' rental markets. An integrated rental market one will find non-profit housing firms aiming at broad parts of the population. Sweden and Austria are examples of such markets. Such countries are likely to have adopted one type of legislation aimed at both the private and the public, or non-profit, housing sectors. A dualistic rental market will on the contrary have a larger degree of separation between the private rental market, and its governing legislation, on the one hand, and the public or non-profit sector. Iceland falls within the dualistic category, with a private rental market and a social housing sector. In this paper, however, we will only consider the private rental market. The reason is that this is the market undergoing structural change, and at the same time facing a discussion of possible changes in the legislation.
From an economics point of view on housing as a commodity, housing differs from other consumption goods in three ways (see Isaac, Allen and Mary, 1991;Quigley, 2002;Kopsch, 2019b). First, housing is a durable good with relatively high production costs and relatively low operating and maintenance costs (see e.g Martin, 2003;Malpezzi, Ozanne and Thibodeau, 1987;Wilhelmsson, 2008; or references within). Second, moving is associated with relatively high transaction costs compared to changing consumption of other goods (see e.g. Piazzesi and Schneider, 2009;Van Ommeren and Van Leuvensteijn, 2005;Hauring and Gill, 2002 or references within). Third, housing is heterogeneous. No house is exactly the same as another, implying there is no perfect substitution to any house.
The first two aspects make housing suitable for renting. Households who fail to meet capital restrictions, because they hold low amounts of wealth or for https://doi.org/10.30672/njsr.95233 other reasons typically turn to the rental market. They cannot pay the large initial production costs, and cannot get loans to cover them. They may however be able to cover the running costs and expenses associated with housing. This will give us a group of households on the rental market without options to leave the rental market.
The rental market may also appeal to some households because of the lower transaction costs. In this group of households we will typically find younger households and students. They have a temporary need for housing and know that they will demand some other type of housing in a relatively near future. Within this group of households one will also find those who have recently moved to a new city. Such households have little information on their own locational preferences, and may choose to rent for the duration it takes them to get acquainted with their new home town.
This distinction of two groups that the rental market can cater to due to attributable differences to housing as a consumption good does not exclude the possibility that other types of households may prefer the rental market, for other reasons. For example, it is likely that we will also find relatively risk averse households on the rental market as renting is associated with a much lower risk of value fluctuations and also because of unforeseen costs, such as for maintenance. We will also likely find households with a strong preference for service provided on the rental market.
The above discussion of the role of rental markets do of course stem from a specific view, housing as a marketized consumption good. To a great extent it is, and under such conditions, one could argue, legislation should be focused on facilitating functioning markets. However, housing can also be viewed as a part of the welfare state, as something that ought to be publicly provided. To some extent, this falls outside of the scope of this paper in terms of the re-distributional aspects that non-profit housing is likely to have on a dualistic rental market. Nevertheless, even private rental market legislation, without re-distribution, is likely to be influenced by a view of housing as a social right. In fact, such a social right point of view may very well be motivated, although not explicitly so, by transaction cost motives.
The view of housing and the role of rental markets is typically described differently between scientific disciplines. The abovementioned economic perspective can very well be complemented by social science perspectives. Noteworthy when discussing rental market legislation is the view of housing as part of the welfare system. This literature can provide us with further insights into understanding rental market legislation. For example, Bengtsson (2001) discusses housing as a social right and what it implies for public policy. Bengtsson argues that the right to housing should be viewed as a political marker, rather than a governmental obligation to supply housing. To some extent, such a political goal might warrant inclusion of housing in traditional welfare provision, along with health care, education and social security. The duality of housing, belonging to both the market and the welfare state has received much attention, starting from Torgersen (1987) and the literature that has followed. Nevertheless, different groups of households will benefit differently from rental market legislation. In this paper we will be interested in three dimensions of rent legislation, namely: security of tenure, legislation restricting rent increases under a contract and restrictions on rent increases between contracts. All these dimension can be viewed from a transactions costs perspective, where stronger legislation in one aspect may lower expected transaction costs to one group, but increase them to another. Let us develop this line of thought.
One of the abovementioned characteristics of housing as a good is that moving, i.e. change in consumption, are associated with relatively high transaction costs. Although transaction costs are relatively lower on the rental market, compared to owning, they are still high enough for households to be likely to demand tenant security, specifically if they have a longer time horizon. Households also make relatively large social investments in their homes. One chooses school for the children, meets new friends in the neighborhood etc. This further increases costs of moving and such investments increase with time. That is, households who are planning to stay long in their rental housing will demand some sort of security in knowing that they can stay for long. That is, legislation providing security of tenure in some way or another is warranted for the rental market to cater to such a group of households. Legislation aimed at tenant security can either take the form of demanding certain lengths of tenure in the contractual agreements, or stipulate under what circumstances landlords or tenants can terminate the contract. If the rental market would only consist of households in transition, security of tenure would be less of a problem. One can think about the market for hotel rooms which solely consists of short term rental, and hence has no legislation providing tenure security.
Legislation providing strong security of tenure demands some sort of restrictions on rents. This brings us to our second dimension of rent legislation, legislation restricting rent increases under a contract (what Arnott, 2003 would call tenancy rent control). Without restrictions to rent increases within contracts, legislation aimed at certain lengths of tenure and prohibiting evictions of tenants may become rather pointless. For example, of legislation would demand a length of tenure of ten years, though stipulating a minimum contractual length, the absence of restrictions on rent increases during those ten years would provide an obvious circumvention of legislation. Landlords could simply increase rents such that renters are forced to move.
Higher tenant security may however come with certain costs. For one, if it is difficult to terminate rental agreements, private individuals may be reluctant to entering them in the first place. Thus the supply of rental apartments will be reduced. In addition, if current tenants receive a high level of tenure security, in terms of length of contracts and protection from rent increases, they will become less likely to move, and thus incurring a cost to those who want to become tenants in the future. Stricter protections to current tenants result in less mobility and a less efficient use of the housing supply (see e.g. Glaeser and Luttmer, 2003). This implies that it will become more difficult to find adequate rental housing. In other words, transaction costs on the rental market increases. The reader may recall that lower transaction costs is one of the benefits of the rental market. https://doi.org/10.30672/njsr.95233 Related to legislation restricting rent increases within a contract for an existing tenant we might also see restrictions on rent increases between contracts and between tenants. Such legislation may be motivated from a policy perspective as ensuring affordable housing, as it implies a general cap on rent levels. Restricting the general rent level may, however, result in other unwarranted outcomes. Rents lower than their market level may for example be counteracted by demands of monetary compensation from landlords when tenants move in, so called key money (Malpezzi, 1988, Kopsch, 2019b. It may also result in tenure conversions where rental apartments are instead sold to owner occupiers (see Diamond et al. 2019;Donner and Kopsch, 2018;Kopsch, 2019b;Turner and Malpezzi, 2003).
Strict legislation in these three dimensions may be beneficial to current tenants, and to those who manage to become tenants. It will be less beneficial to landlords, who will face larger difficulties in increasing rents or choosing their tenants. In addition, by reducing supply both through reducing the frequency with which households move and by shifting incentives toward supplying apartments on the owner occupant market, strict legislation may hamper the possibilities for a household to become a renter. Legislation has to balance the interests of these three groups, landlords, tenants, and those who want to become tenants.
The further discussion of the legislation governing the Icelandic rental market will focus on these three dimensions; security of tenure, restrictions on rent increases under a contract and restrictions on rent increases between contracts. An interesting aspect is that the structure of the rental market, e.g. what type of households reside there, for how long and for what reasons, as well as what type of financial actors enter the market, and the legislation will not develop together in an organic fashion. Legislation will have to adapt to changes in the rental market. We will first turn to how the Icelandic rental market has changed over time.

Evidence of structural change under the current rental legislation
In the following section we rely on data gathered by Statistics Iceland for the EU-SILC survey. These surveys have been conducted since 2004. In addition, we rely on national surveys of the housing market, including renters, made in the years 2003 and 2015 by Gallup for the Ministry of Social Affairs 1 , 2007 by University of Iceland, Social Science Research Institute and since 2017 by Zenter research for the Housing and Construction Authority in Iceland and its predecessor. 2 The included questions vary over the years which makes statistical testing rather difficult. Instead, we will here present our argument in a more qualitative fashion, looking at what we will argue to be evidence of structural changes.
We argue that the gathered evidence from the available surveys allow for the conclusion that the Icelandic rental market has been undergoing structural change. We will in the following present seven distinct arguments in favor of 1 The Ministry of Social Affairs has undergone name changes. In 2015 it was called the Ministry of Welfare. 2 In the beginning of 2020, its predecessor the Housing Financing Fund and the Iceland Construction Authority merged to form the new institution, Housing and Construction Authority. such a structural change, along with how these structural changes may or may not warrant changes to legislation.
First, since the financial crisis both supply and demand on the rental market has increased. After the collapse of the banking system a lot of newly built and half-finished housing was available. House prices fell and were expected to fall even more. Due to an increased uncertainty of real estate prices, decreased income, decreased access to capital excluding certain households from buying and negative equity of households it was hard to sell real estate. Many decided to rent out their properties until conditions would improve on the market (Institute of Economic Studies, 2011). Between 2007 and 2012 the share of Icelandic households residing on the rental market increased from 17,4% to 28,3% according to EU-SILC survey gathered from Statistics Iceland. That is, an increase of 70% in the number of homes. Figure 1 depicts the share of households who reside on the rental market in Iceland during the period 2004 to 2016. 3 The share includes both the general rental market, i.e. those who rent at market rate, as well as those who rent at reduced rate such as social housing, student housing and other types of rentals. In the immediate years after the financial crisis the general rental market grew the most. In 2012 the number of households on the private rental market were 50% higher than the number of households renting at reduced rate. However, in 2016 the size of these markets have converged to equal size. From figure 1 one can make the argument that the rental market has stabilized at a new, larger equilibrium. A larger rental market is not by itself an argument to change the governing legislation. We must in addition gain some understanding as to why the rental market has been increasing in size over the past years.  These results apply to the total rental market. If we instead look only at the private rental market (from here on excluding friends and family unless stated otherwise), the difference becomes even larger. In 2019 around 72% of private market renters were renters by necessity, 19% saw it as temporary and a mere 10% had a preference for renting. This can be compared to the 2003 survey, where 46% of private market renters stated necessity, 16% by preference and 38% saw renting as a temporary solution.
Thus, there is evidence that the rental market is growing partly because some households do no longer see any other options. Such households may demand tenure security to a larger extent than what renters have demanded in the past, when the rental market to a larger extent catered to temporary housing needs. If so, this change can constitute part of an argument for more protection through legislation on the rental market, which will demand changes to existing legislation. However, as we will turn to now, tenant security may improve even without legislation.
Third, the length of tenure has increased, i.e. tenants report that they stay longer on average in their current rental apartments than in the years before the financial crisis. Figure 2 shows that there is a clear change in the share of tenants on the total rental market with tenure less than a year from 2003 compared to the more recent years. We can also see an increase in the share of renters with longer tenure. This is consistent with the hypothesis that the growing over all rental market can be explained, at least partly, by increasing difficulties for households to purchase their own housing and that the share of temporary renters are decreasing.
In 2003 the average length of tenure was 2.5 years. By 2015 this average had increased to 2.8 years in 2015 and measured at 2.6 in 2019. The averages do not imply much of a change. If we instead look at the development on the private rental market alone, the average length of tenure was 1.4 years in 2003, 2.2 years in 2015, 2.1 years in 2018 and 1.9 in 2019. That implies roughly a 48% increase in the later survey years compared to the pre-financial crisis market. 5 4 The survey conducted in 2007 had slightly different questions and was quite smaller in scale regarding the number of renters who answered. It had the same options as in the other surveys but included an option: "temporary and necessity". This option captures renters who rent temporarily due to necessity reasons and thus these answers were combined with the temporary option by the authors of this paper. Fourth, there has been a rise, or rather an awakening, of large professional for-profit rental companies. Rental companies did exist before the crisis but they were very small and didn't have a large market share. Their exclusion from the early surveys is telling to this fact. Comparing survey data from 2003 to recent years we can see how the structure of the market has changed, figure 3 provides a visualization.
The most evident change is the rise of the professional rental company as well as the decreasing share of municipal social housing and of the private individual arm's length landlord. The share of those renting from friends and family has more or less stayed the same, with an average of 20%   share of around 14-16% in recent surveys whereof two companies have a total share of around 8-9% 7 or about 60% of the share of these companies. This might not seem very much but this is the share of the total rental market which includes social and student housing. If we only look at the private market and exclude other types of housing that are rented out at a reduced rate and are not available to all, i.e. social housing, student housing and housing from non-profits, the market share of these companies could have been around 20% during that period and as high as 30% in 2018 and 2019 if we also exclude rentals from friends and family which are typically rented out below market rate.
The rise of the private rental companies can be partly explained by the development of residential property and rental prices after the financial crisis. In the years after the crisis residential housing was believed to be undervalued which attracted investors. In January 2011, real prices in the capital region had fallen by 34-37% from the peak in January 2008, depending on whether the CPI including or excluding housing costs is used for deflation, see figure 4.
In 2014 two companies were formed that in the coming years would merge with smaller rental companies and buy up large stocks of housing available, for example from the Housing Financing Fund, and other apartments already on the rental market.
The increasing share of professional rental companies has brought a few changes to the rental market. One such change is that more households now have a written contract than before. 87% renting from arm's length individuals on the private market answer affirmative to the question whether they had a written rental contract in the beginning of the tenancy or not, compared to 97%-98% of those renting from the professional companies. In 2003 around 80% on the private 7 This is a rough estimate using the total number of rental apartments reported in annual statements of the two largest companies divided by an estimate of the number of rentals on the market, using Statistic Iceland's most recent number. market had a written contract. When asked whether they had a written contract at the time of question, the percentages in the surveys are very similar. Having a notarized rental contract adds security to the tenant and is a prerequisite to be able to apply for rental allowance, although not needed in the case for social and student housing since 2017.
Professional companies are more likely to initially make short-term contracts than individual landlords on the private market. In 2019, 64% of tenants renting from professional companies had a fixed-term contract between 12-23 months made in the beginning as compared to 51% of those renting from arm's length individuals on the private market. Only 13% had made a non-fixed term contract with the professional companies compared to 27% of those renting from arm's length individuals on the private market.
Although initially shorter, rental contracts with professional companies are renewed to a higher degree than contracts made with private arm's length individuals. The reasons may be several. For example, tenants may sort due to socioeconomic status, with more affluent tenants sorting to professional companies than the traditional market. There are signs in recent surveys that it might be more common for households with higher income to rent from these companies than households in the lower income brackets, maybe due to selection of the companies themselves. Another explanation can be found in that professional rental companies are less likely to demand the apartment for personal needs, than a private individual. The sole purpose of ownership of an apartment by a professional company is to rent it out long term to reduce transaction costs. However, it is not clear if all of these companies are in it for the long term and might as well sell off their portfolio when the price is right.
In 2019, 70% of those renting from professional companies had their contract renewed as compared to 44% renting from arm's length individuals on the private market. This can probably explain some of the difference in perceived tenant security of those renting from these companies as compared to those renting from private arm's length individuals. Tenant security on the private market has stayed more or less the same in recent years but has decreased compared to 2003. In 2018, 13% renting from professional companies felt that it was rather likely or very likely that they would lose their apartment as compared to 32% of those renting from arm's length individuals on the private market. These results are almost identical in the 2015 survey. In 2003, 22% of tenants on the private market felt it was rather likely or very likely to lose their apartment. To compare this with 2018 and 2015, the weighted average for the private market is 26% and 27% respectively, when taking into account the professional companies. The survey conducted in 2019 can be interpreted to show similar results as in 2018 but it contained a slightly different question and thus cannot be directly compared to the earlier surveys. 8 The rise of professional companies thus seems to have mitigated the decreasing tenant security on the private market. However, according to the 2019 survey, 91% of those renting from professional companies who had their initial contract renewed were offered a new contract with increased rent. The corresponding share for renters with arm's length individuals as landlords is 51%. There is also a difference in perceived bargaining power. In 2019, 46% of tenants renting from arm's length individuals on the private market considered their bargaining power to be rather strong or very strong and 26% of them rather weak or very weak. Contrasting to this, only 25% of tenants renting from professional companies perceived their bargaining power to be rather strong or very strong and 53% stated their bargaining power was rather weak or very weak.
Between January 2014 and January 2020, the price of rent in the capital region increased by 59% according to Registers Iceland rental price index. If deflated with the CPI, real rent has increased by 40% in the same period. This can probably be explained by the housing shortage that developed in the period due to lack of new housing constructions in the years after the financial crisis along with immigration effects and negative supply shock of long-term rentals due to Airbnb following the tourism boom, where the number of tourists almost quintupled between 2010 and the peak year 2018.
On the one hand, the increase of the market share of professional rental companies may demand according changes to legislation, clarifying tenant rights in tenure security and rent changes. From the perspective of the legislator it may be relevant to distinguish private landlords from each other. In particular, one might want to distinguish those who own apartments in the sole purpose of renting them out from those who do so for a transitional period when the personal 8 Tenants were asked how much they agreed or disagreed with the statement that they believed they had tenant security. Respondents renting from arm's length individuals answered in the following way: 18% strongly agreed, 35% rather agreed, 7% neither/nor, 11% rather disagreed and 15% strongly disagreed. Those renting from professional companies were a little more positive or: 18% strongly agreed, 44% rather agreed, 17% neither/nor, 11% rather disagreed and 8% strongly disagreed.
need of the apartment is absent. On the other hand, the empirical observation is that professional rental companies, even in absence of legislation, offer a larger perceived tenant security than the traditional rental market. 9 Tenant security, realized in fewer moves by renters, is what we turn to next.
Fifth, the number of times people on the rental market have moved is trending downwards. In 2003 the average number of moves was 4.3 in a 10 year period. In 2019 this average had decreased to 3.6 moves. The same trend is visible for other time frames, 2.9 moves in the last 6 years in 2003 compared to 2.6 moves in 2019 and 1.8 moves in the last 3 years in 2003 compared to 1.5 moves in 2019. Regardless of chosen time frame, the average number of moves for households on the rental market have been declining over the past two decades, as depicted by figure 5. In 2019, the number of moves for those renting from a professional company was significantly lower than the number of moves of those renting from private arm's length individuals for both the past six years and three years.
The decreasing number of moves is in line with the observation of increasing length of tenure. These two observations, decrease in moves and increase in length of tenure, are interesting, especially when taking into account the rapid rise of Airbnb short-term rentals from 2014 to 2018. Mermet (2019) finds evidence of direct displacement where the landlords cancel rental contracts in order to turn the property to an Airbnb-rental. Elíasson and Ragnarsson (2018) estimated 9 There were some changes made to the legislation in 2016 but only minor changes which are relevant to this discussion. Notice period for the termination of a rental contract for single rooms was increased from one month to three months for both parties. The notice period that for-profit rental companies must give their tenants when terminating a contract was increased from 6 months to 12 months, if they have lived in the apartment for a year. The longer notice period a landlord must give tenants who have lived in the same rental for 5 years used to be 12 months but was abolished. It is now the same regardlesss of length of tenure, 6 months. This, however, hasn't affected the perceived tenant security as the results in 2015 and 2018 are almost identical. that around 1700 apartments from the residential housing stock had been taken over by Airbnb activity by the end of 2017, 1200 thereof in the capital region, if measured as apartments for more than 150 nights during a twelve month period. This amounts to around 10% of the estimated number of homes on the general rental market at the time. Legislation with the aim of restricting Airbnb-rentals was implemented in Iceland in 2017.
These past observations, regarding the increased length of tenure, decreased number of moves and increase of professional rental companies all speak to the fact that tenant security can increase without much changes to legislation. However, tenants on the private market might not feel that way as perceived tenant security has decreased a bit since before the crisis. One should also recall that we are looking at averages, which can of course hide individual cases where such legislation could have been fruitful. In addition, observations are not unambiguously pointing at improvements for tenants. As mentioned, while offering greater possibilities to keep on renting, professional companies are charging higher rents in renewal. Again, this speaks in favor of a possible duality in legislation, differing between the intention or identity of the landlord.
Sixth, from 2003 there has been a substantial change in how likely tenants believe it to be they will stay on the rental market. The share of renters stating it to be likely they are still on the rental market in a year has increased from 72.5% in 2003 to 89.3% in 2019. Households who believe it be unlikely or almost unthinkable that they are still on the rental market within a year has, in the same time period, decreased from 21.8% to 8.2%. These results reflect the total rental market. When we look at the private market only the differences are even larger. The share of households who believe it to be almost certain or likely has increased from 66% in 2003 to 90% in 2019, and those stating it to be almost unthinkable or unlikely has decreased from 27% to 8%.
When asked about longer time horizons the trend is the same but the difference becomes even clearer. These differences are depicted in figure 6. Panel A present the results for the rental market as a whole, including all types of rentals. It can clearly be seen that households expect to stay on the rental market to a greater extent in the 2019 survey than in the 2003 survey. Panel B present the results broken down for the private rental market only. The picture becomes even clearer, and we might hypothesize that the temporal changes are driven primarily by this market.
What has been presented here could be an indicator that the market is stabilizing as a long-term solution for a larger group of households. The explanation could entail that they are trapped on the rental market, i.e. they cannot for some reason enter the ownership market. That would mean we have a larger group of households who do not qualify for subsidized housing, but at the same time cannot enter the ownership market. This argument gains further strength if we look at the fact that it has become increasingly more difficult to save the own equity required to buy an apartment. This difficulty is likely driven both by the increasing burden of housing cost for renters and increasing prices of housing.
Panel A of figure 7 depicts the share of households on the rental market and homeowners living with housing costs that exceed the overburden rate. The overburden rate is the share of housing costs of 40% or more of disposable income, as defined by Eurostat. Panel B depicts the median housing cost burden borne by households. The housing cost burden is the average share of disposable income spent on housing. The share of households on the rental market with housing costs of 40% or more of their disposable income increased from 12.6% in 2007 to 26.4% in 2015. However, in recent years the cost burden has stabilized, or even trended downwards due to a rapid rise in purchasing power. According to Statistics Iceland's income accounts, between 2007 and 2015, purchasing power of disposable income per capita declined by 11.5%, from 2015 to 2018 it increased by 13.2%.
That more households have fewer options is an obvious case in favor of changing legislation. As we will get into in the next section, having the option to move is a factor that affects the bargaining power between renters and landlords. One purpose of legislation on the rental market is to balance such bargaining powers.
Seventh, the types of households who reside on the rental market have changed. For example, the share of households consisting of a single parent has changed the most. Comparing averages between the two periods 2004 through 2008 on the one hand and 2012 through 2016 on the other, the share of single parent households residing on the market has increased from 32% to 50%, meaning half of such households are renters. The number of two-parent households with children is also increasing, from 9% in the pre-financial crisis years to 17% in the later period.
There is also a difference between the genders. The share of households consisting of a single adult male have increased by over 14 percentage points, from 35% to 50%, while the share of households consisting of a single adult female has slightly increased from 26% to 30%.
From 2004 to 2016 there is also a clear increase in the share of individuals in the lower income brackets who reside on the rental market, see figure 8. In 2004, 24.9% of individuals with income in the lowest 20% of the income distribution were on the rental market. By 2016 this share had increased to 44.2%. This is for the total rental market, however if only the general rental market is taken into consideration, there is a 98% increase in the number of individuals in the lowest 20% of the income distribution who reside on the general rental market, from 9.6% in 2004 to 19% in 2016. Summing up, there is vast evidence of an Icelandic rental market undergoing structural change. Most notably from a perspective of legislation may be the rise of professional companies devoted at renting out apartments, coupled with an increase in the number of households, and types of households, and the share of lower income individuals residing on the rental market. Such changes may warrant corresponding changes in legislation.

The Icelandic legislation and possible motives for changes
In the following section we will provide a brief description of the current legislation on the Icelandic rental market, Rent Act, No. 36/1994, in the three dimensions previously described. 10 The discussion will be paired with what changes may be motivated, taking the previous discussions of the role of rental markets, and the argued structural changes of the Icelandic rental market, into account.

Tenant security
In Iceland, rental agreements can either be fixed-term or valid until further notice ( §9). A contract is regarded as valid until further notice unless otherwise stated. According to §10, if a written contract has not been made, the agreement is regarded to be valid until further notice. The rent is then the amount that the landlord can prove that the tenant has agreed upon.
As stated in §55, both parties can terminate a contract valid until further notice. Termination should be done in writing and sent in a verifiable manner. According to §56, the notice period for termination of contracts of this type is equal for both parties, six months. However, if the tenant has lived in the apartment for over twelve months and is renting from a professional for-profit company then 10 The Rent Act, No. 36/1994was ammended by Act No. 65/2006, Act No. 66/2010, Act No. 162/2010, Act No. 77/2011, Act No. 126/2011and Act No. 63/2016  the landlord must give a twelve month's notice. The notice period for single rooms is three months for both parties. Fixed-term or temporary contracts end at the agreed date and cannot be terminated during the time of contract unless certain pre-agreed circumstances apply which have to be stated clearly in the contract ( §58). In which case both the tenant and the landlord have a three month's notice. There is no minimum period for temporary contracts, however, in the case of a tenant renting from his employer the period of tenancy cannot be less than the period of employment ( §50).
According to §59, if eight weeks pass from the point the tenancy ends according to the termination of either a contract valid until further notice or fixedterm contract, but the tenant continues renting the residence and otherwise fulfills the requirements of the contract, the contract will be automatically renewed as a contract valid until further notice, as long as the landlord hasn't specifically asked the tenant to vacate the property after the contract ended.
A landlord can terminate a contract without notice under certain circumstances stated in §61, for example if the tenant fails to pay rent for some time or due to other breaches of contract. The tenant has also the right to do the same if the landlord for example hasn't fulfilled his duties regarding the rental property ( §60).
During 2020 discussions has been ongoing regarding changes in tenant security. Among discussed changes are the possibilities for landlords to terminate rental contracts. In practice, the law will demand landlords to show reason for termination of contracts valid until further notice. Among the suggestions are that landlords can only terminate a rental agreement if the rented property is within the same premises as the landlord resides, if the apartment is rented out furnished or if the landlord needs the apartment for herself, a relative or if she intends to sell it or carry out large renovations.
Moving towards a stronger tenant security may very well be warranted given the above presented information of changes on the rental market. We have argued that rental legislation should cater to both households who see the rental market as a transitional solution and to those who see it as a long-term solution. The latter group has been growing in Iceland. More households are on the rental market, they stay longer and they expect do so in the future. At the same time, proposing possibilities for the landlord to terminate a contract to use the premises herself ensures a difference between private landlords and professional for-profit rental companies, as the latter will have difficulties arguing for the need of a specific apartment.

Changes to rents under contract
Icelandic legislation gives a high degree of freedom of contract both regarding the initial rent and rent increases under contract. According to §37, the landlord and the tenant are free to negotiate how the rent changes during the time of contract. It is, however, common practice in Iceland that rent under contract follow changes in the consumer price index. However, the law requires the rent to be considered fair and reasonable with respect to both parties and the housing complaints committee can decide whether it is or not.
The definition of fair is rather vague. According to explanations given in the statement with the original law proposal of the current rental legislation what should be considered fair should always depend on location, circumstances and the current state of the rental market. 11 The main reference should be the market rate of comparable rentals but other factors can be taken into account, such as: operating costs, interest expenses, taxes, type and condition of the property and cost of maintenance. A complaint can be filed with the housing complaints committee if for example the tenant wants to dispute the rent increase.
In contracts valid until further notice, conditions for rent changes have to be clearly stated at the beginning. The landlord can however terminate such a contract, with appropriate notice period, to renegotiate the rent. The same applies with fixedterm contracts except they cannot be terminated before they expire unless it is clearly stated in the contract under which circumstances they can be terminated. This latter possibility, termination of rental agreements in order to increase rents is an aspect that may warrant changes, specifically if the legislator wants to move in a direction of larger tenant security. As has been discussed above, there is a case to restrict possibilities to terminate rental agreements from the landlord side. The current legislation, that leads to incentives to terminate contracts to increase rents can however be taken as evidence that such problems may be solved by restricting termination possibilities. However, the arguments for structural change of the Icelandic rental market shows that new types of households enter the market, specifically households with children. At the same time, households fail to see any options to renting. This implies that there might be a case for restrictions of rent increases within contracts as well. This is a difficult topic though, as such restrictions will become a re-distributional policy, and such policies have other substitutes that may be more accurate in terms of what households receive help. For example, the legislator could choose to not restricts rents within contracts and instead help households with housing allowances or subsidize construction of rental housing for lower income households.

Changes to rents between contracts
There are not many restrictions on rents for new contracts in Iceland. As with changes to rents within contracts, the same limitation holds, rents should be fair ( §37 and §53). When a contract expires and the landlord wishes to offer the apartment for rent again (for at least a year), the tenant has priority ( §51). However, under certain circumstances which are clearly stated in the legislation the priority does not apply, for example if the landlord intends to sell the apartment within 6 months after the contract expires or make it available for relatives.
When a contract is renewed the conditions of the new contract should be the same as in the original as long as those conditions can be considered fair. The law assumes that it is likely that the level of rent in the original contract is fair and the party that disputes the current level has the burden of proof. https://doi.org/10.30672/njsr.95233

Discussion and conclusions
With this paper, we have aimed at discussing the legislation governing the Icelandic rental market. In particular, we have been interested in structural changes to the Icelandic rental market and how such changes may warrant changes to the current legislation. It should be stressed that the conclusions we will draw are based on the historical changes of the rental market up to date. Changing economic conditions, such as globally falling interest rates or increased unemployment in Iceland, may warrant other conclusions.
In our argumentation of a structurally changing rental market in Iceland we put particular emphasis on certain changes. For one, the traditional Icelandic rental market based on individual households subletting a second home has begun to transform into a market where professional rental companies plays a larger role. Such companies do offer tenants greater security when it comes to length of tenure, but they offer on average less security in terms of changes to rents within tenure. Since more households are on the rental market for longer, with the expectancy of staying there, it may be warranted to implement corresponding changes to the Icelandic legislation of tenant security.
In terms of transaction costs, there is growing tendency for households to stay on the rental market for a longer period of time. This implies that there is a possible argument to protect them from the necessity to move, thus restricting landlord's possibilities to terminate contracts. There is also a need for a discussion of restrictions to rent increases within contracts. This can be motivated partly as way of ensuring that tenant security becomes stronger. It may also be motivated by changes in what types of households reside on the rental market. There has been a shift resulting in families with children, particularly single-parent households, residing on the rental market to a larger extent than previously. There is also a shift in perceived options, households on the rental market see fewer options than before. This implies there might be more households on the rental market today in need of financial help. Such financial help can be provided through legislation directed at restricting rents, or through monetary re-distribution via housing allowances.
In summary. The changing structure of the Icelandic rental market can at least be argued to demand changes in tenant security. It may however be warranted to differentiate between types of landlords, where landlords who are private individuals with one extra home are more likely to exit the market if restrictions become too harsh, compared to private for profit companies who can bear a larger financial risk. However, it is possible that some companies might sell their properties if legislation becomes unfavorable.