The model of tenure choice in life cycle of Poles – empirical analysis

: We analyse tenure choice determinants of Poles for the years 2006, 2010, 2014 and 2018 on EU-SILC data, distinguishing between co-residence with the parents, renting at market rates and owning a house. The average marginal effects obtained from multinomial logistic regressions allows us to conclude that ownership is strongly connected to the family status, but this relationship weakens over time. Unmarried people without children but with better incomes become increas-ingly more likely to be homeowners. Thus, ownership is not anymore a main domain of conventional families. Moreover, renting becomes an option to become independent for those who cannot afford to buy a house. But it is rather a temporary solution, because renting is more costly than buying a house with a mortgage. We also find evidence of the gender gap in homeownership, as women are more likely to be owners, and less likely to co-reside with their parents than men. Along the evolution of the housing finance system, growing wages make ownership more likely. We observe that the choices are more free in densely populated regions, where the rental market develops. In this part of the paper explain the data that we use, present the empirical results and discuss them.

The model of tenure choice in life cycle of Poles -empirical analysis to inherit (making their status closer to that of a tenant). In consequence, this group is very heterogeneous (this problem has been identified by Matel, 2020). On account of the state's withdrawal from financing the social-housing market the number of people who live rent-free in such accommodations is falling -these flats are either sold to private owners or fall into disuse on account of their poor conditions. The authors have, therefore, decided to present free accommodation in social housing and free accommodation in apartments belonging to private owners as a category of tenure but without any analyses of its determinants.
Based on an analysis of literature, analysis of a relationship with a dependent variable, and analysis of a relationship between independent variables, as well as a diagnosis of collinearity we have identified a set of eight independent variables. Four of these were nominal, two were ordinal and two quantitative. Additionally, since their housing status did not vary, we have decided to exclude widowed people from the analysis (1,7% tenants 1,6% -living with parents w 2018 r.). We have decided to combine workers and retirees since the current housing status of the latter is to a large degree a consequence of their past professional situation. Degree of urbanization § densely-populated area -areas with population density of no less than 1500 people per km 2 , and town population of no less than 50,000 § intermediate area -areas with population density of no less than 30 people per square kilometre, and town population of no less than 5,000 § thinly-populated area -all other areas (base) Ordinal Source: own elaboration: Methodological guidelines and description of EU-SILC target variables, European Commission, Eurostat 2006, 2010, 2014, 2018 In our analyses, we made use of multinomial logistic regressions. Four models illustrating the current tenure status of individuals (for 2006, 2010, 2014, and 2018) were constructed. The model can be expressed as follows (Agresti, 2013): where π i is the probability of living with parents/owning/renting for the individual i and X i (X 1i ,X 2i , X 3i …) is the vector of individual level covariates.
The regression results are presented in two stages. The first stage are tables with the usual regression coefficients, while the second stage is table with the average marginal effects (AMEs). We focus especially on the latter, which allows for a ceteris paribus analysis of the change of the probability of the dependent variable, when the remaining control variables are held constant (Williams, 2012). The AMEs can be directly compared across groups and between models (Fiori et al., 2019). Note: Estimates include sample survey weights.
Source: EU-SILC databases 2006, 2010, 2014, 2018 During the period covered by analysis the percentage of homeowners made up of people who had never been married and divorcees increased. In general, the percentage of single people who lived with their parents fell although in 2010 it had experienced a resurgence. This, in our opinion, was connected with the country's period of economic recession. On the one hand, during a time of an economic depression, people who have never been married have a more difficult time in overcoming income barriers to become homeowners while on the other continuing to live with their parents becomes more acceptable. During the considered time the structure of tenants changed, a fact that was mainly The model of tenure choice in life cycle of Poles -empirical analysis connected with the overall rise in the population of never-married and divorced people between 2006 and 2010. The percentage of women is greater among homeowners and tenants than among people co-residing with their parents, a result of the fact that women generally become independent earlier. In considering age it can be seen that renting is a transitional form of accommodation between living with one's parents and homeownership. Interestingly, in Poland, the difference between the income of a tenant and a homeowner is dwindling. This is connected to the fact that renting in Poland is as costly or more expensive than owning a home. There are also considerable differences in the tenure status of people that are caused by variations in the level of urban development. Rental properties exist mainly in large cities while smaller municipalities are characterized by a dominance of single-family housing which creates greater opportunity for living with one's parents. Table 4 while average marginal effects obtained from a multinomial logistic regression that shows how a percentage change of a given variable affects the probability of a tenure choice are shown in table 5. Our research indicates that housing choices in Poland are in some way typical for particular age cohorts. Poles first live with their parents and then seek independence most often aiming at homeownership. Some go through the transitional stage of renting. In Poland, this cycle has not significantly shifted over time and is consistent with the concept of the housing ladder according to which the first step, that of purchasing one's first, small flat, is the most difficult. After that, it is only necessary to add a marginal but still very important amount to the money obtained from the sale of the first flat to get a better or bigger one. At the same time, it is possible to see some differences within this process between women and men. In Poland women, on average, leave the home of their parents and get married, an event that in Poland is strongly connected with the decision to purchase a home, earlier than men. This results in a higher probability for women to become homeowners and a smaller likelihood of their co-residing with parents with the first difference persisting over time while the second grows. For men, on the other hand, the probability that they will continue living with their parents longer increases. Generally in Poland, an increase in the share of young adults aged 25-34 living with their parents has been observed (in 2006 it was 37.8% while in 2018 it reached 45.1%, Eurostat) and concerns men more than it does women (in 2018 it was 52.6% for men in comparison to 37.3% for women). At the same time, no differences with respect to renting have been seen (see Maroto and Severson, 2020) which, in turn, means that women move out faster but only to their own home while renting is equally popular among women as it is among men (see Matel, Olszewski, 2021).     Notes: * p <.05, ** p <.01, *** <.001, estimates weighted using respondents weight from the survey. Source: EU-SILC Poland, 2006, 2010, 2014, 2018 A key role in the shaping of housing choices in Poland is played by a household's life cycle. It can be seen that marital status has a significant (outright dominant) influence on determining the living arrangement of Poles (similar for Scots, Fiori et al., 2019, for the Dutch -Feijten, 2005; and the British - Thomas, Mulder, 2016), but, at the same time, a decrease of its importance over time can also be observed. People who have never been married and divorced people, in 2018 exhibited nearly 30% and 18%, respectively, smaller probability of homeownership than married people (this is similar to the Dutch, see Blaauboer, 2010), while in 2006 this difference exceeded 41% in the never-married group and 27% in the divorced group. Singles, meanwhile, demonstrated a greater likelihood of living with their parents, and the differences were larger for never-married than divorced people and fell over time for both groups (similar results were obtained by Mundra and Uwaifo-Oyelere, 2016).

Multilogit regression results have been presented in
The present study confirms that the rental market to a greater degree fulfills the needs of single people and with its development, the probability of living in rented housing rises for divorcees as well as, although to a lesser degree, those who had never been married. Generally, however, it is possible to observe that differences between never-married and married people are smaller than those between divorced and married people but the trajectory is the same. Housing models of divorcees resemble more those of people who had never been married than married people since a divorce fundamentally changes the living arrangements of both partners. The connection between housing choices and the life cycle is also reflected in the importance of having children. People who have kids display a higher probability of homeownership and live with their parents less often. This, however, has no impact on renting. The influence of having children on living with one's parents falls over time. Generally, it is possible to see that in Poland the decision to become independent is less and less relevant with respect to starting a family.
Within the last several years the significance of higher education with respect to homeownership is also falling in Poland, although those who had such, in 2018 displayed a 4% higher likelihood of living in their own home and a 2.5% less probability of living with their parents. Since 2018 education became a new source of inequality in the tenant group. People possessing higher education presented a 2% smaller probability of living in rented housing. This could result from the greater occupational stability of people working in professions requiring a university degree which translates to better access to home loans. Generally, however, it can be noted that this variable loses its significance concerning the overall household income. Along with a rise of income expressed in 1,000 Euro per person per month within a household, the probability of homeownership grows and this influence intensifies over time. Furthermore, in 2006 it was not the level of income but factors connected to family dynamics that decided about ownership. In our opinion, this was the result of the overall shortage of housing The model of tenure choice in life cycle of Poles -empirical analysis as well as the fact that homeownership was often the only alternative to become independent for people who started a family or ones who experienced greater pressure, both external and internal, to gain independence. After 2006 the substantial rise in home prices, increased demand, and intensified promotion of loans caused income-related disproportions in access to homeownership. In effect, in 2006 a decision to become independent was not connected to one's level of income. However, in 2010 it was possible to observe a growing influence of income on independence but only in relation to ownership. In Poland renting is, in reality, as costly as homeownership. So, living in rented housing is, to a great extent, a question of lifestyle and usually, this is a transitional stage between living with one's parents and homeownership. However, people who can afford to rent most often can also afford to buy a home, the reason no income inequality can be seen between homeowners and tenants. In Poland, therefore, people who rent are not those who earn too little but rather those who are at a particular stage in their life during which renting is preferred over ownership.
When it comes to Poland, the variable that is key in understanding the decision to rent is urban development. The rental market has only started expanding recently and that growth has not been balanced. It has been concentrated in large cities which is connected both to housing demand as well as supply. In large urban centers, especially university towns, landlords have additional prospective tenants -university students, a large group of consumers within the rental market. In effect, the likelihood of renting is significantly higher in densely populated than in thinly populated areas and these differences grow with time along with the development of the renting market. Within the last several years these differences (although smaller) have also been observed between intermediate and thinly populated regions. In reality, therefore, for people who live in large cities, especially those who have migrated from less urbanized areas, housing choices are limited. Since they had migrated to a new place they cannot live with their parents which leaves them with the choices of renting or buying a home. In turn, people who inhabit smaller towns also encounter the problem of diminished choices and must choose between living with their parents or ownership.

Discussion and conclusions
Poland is a country that is dominated by homeownership and a sluggishly developing renting market whose expansion has been significantly slowed by the economic recession caused by the global financial crisis. Like other post-transformation countries, the state's engagement in social housing is diminishing while housing subsidies are aimed mainly at those who want to become homeowners. For the last several years the development of the new construction market and a rise in incomes have resulted in increased interest in privately-owned housing. On the other hand, low interest, relaxation of tenant protection (Łaszek et al., 2021) as well as filtration processes releasing some assets making them available for rent (Brzezicka et al., 2019) have slowly stimulated the development of the renting market even though the costs of renting are still higher than those of buying a home with a mortgage loan. In effect, only people who treat renting as something transitional, those who want to remain mobile or those who are forced to become independent and cannot overcome barriers preventing them from becoming homeowners decide to become tenants (Rubaszek, Czerniak, 2017).
Our research indicates that the housing choices of Polish people during the period of intense market change are altering but this transition is very slow. This is accompanied by a change in both the scope as well as the importance of factors that shape them. Interestingly, in the years 2006-2018, the direction of changes in the impact of most of them is the same, regardless of the phase of the business cycle. This, in turn, indicates that the changes in the housing choices of Poles are strongly related to housing system transformation and its development and only to a small degree on the business cycle. In our opinion, this is the result of several mechanisms. First, as has been mentioned in part 3 of the present article, the effects of the recession following the GFC were not as severe as those experienced by numerous other countries and wages continued to grow. Second, changes seen in the structure of the housing status of Poles during the recession (2010) are, in reality, the materialization of pre-sale contracts signed between 2006 and 2008. The impact of institutional changes on the housing status of Polish people can be seen allowing the explanation for the strong growth of the rental market even though it remains quite insignificant (4.5%). Another important cause for observed changes concerns alterations in the life cycle of Polish families with independence and the moment of starting a family coming later and later.
Generally our studies confirm that housing choices are strongly connected with the life cycle (similarly to, among others, Clark, Mulder, 2000, Blaauber, 2010Barrios et al., 2013, Xhignesse et al., 2014, Špalková, Špalek, 2014, Fiori et al., 2019, Mundra, Uwaifo Oyelere, 2019, Maroto, Severson, 2020. Entering into and the dissolution of marriage have the greatest impact on the housing choices of Poles during the entire duration of the considered period although this does diminish over time. In our opinion, this is related to the additional option of becoming independent, that of renting. This decrease in influence also means that homeownership is no longer reserved only for married couples. The market is observing the formation of a group of consumers who have never been married and who are interested in buying a home before formalizing their relationship. We believe that this is a new, important group of consumers that may have different housing needs. Furthermore, study results have shown that unmarried people not only buy homes more often but more often move to rented housing which means that, generally, their decision is to become independent. This, in turn, signifies The model of tenure choice in life cycle of Poles -empirical analysis that Poles who co-resided with their parents more and more frequently move out to live on their own before they get married and start a family. This is connected to growing incomes, accessibility of home loans for people not within formal relationships as well as socio-cultural changes with young Poles generally getting married later and postponing having children. Moreover, greater availability of housing, growing competition on the new housing market, and intensive housing construction allows the delivery of homes fulfilling not only the needs of families. This especially concerns accommodations that are smaller, less expensive, or located in places near many amenities such as schools and workplaces. Our studies have also shown that the developing renting sector is an especially attractive alternative for divorcees for whom moving back in with their parents means the necessity of coming back to their family home which may be more difficult than for single people who simply have not yet become independent. Our research has confirmed that in Poland there is a certain gender homeownership gap which turned out not to concern differences between renting and homeownership but rather those between ownership and co-residence (similar conclusions were reached for Canada by Maroto, Severson, 2020). In Poland, like in Canada, women leave their family home as well as get married sooner than men and, in effect, more often become homeowners. In reality, at current prices in Poland renting is an alternative for people who want to become independent but for reasons of their own do not want to buy a home (need to retain their occupational mobility, imprecise preferences concerning a home, expect changes in their family situation) or who are at the stage of gathering funds to fulfill this goal (which justifies a lack of differences in the levels of income of homeowners and tenants in Poland).
An interesting observation is a fall in the significance of having a higher level of education in gaining access to homeownership in favor of income level. This concerns social changes with the higher level of education of the general population causing shortages of skilled technical workers and growth of average wages of support professionals not requiring a university degree 1 . In Poland, it is possible to observe income inequalities between people who co-reside with their parents and homeowners which seems to be intuitive. No inequalities exist, however, between tenants and homeowners. In Poland renting is an attractive option not on account of the current price of maintenance but rather other characteristics, primarily flexibility and mobility, and is an attractive alternative for people who have not settled down yet both concerning the family as well as professionally, especially for those whose occupational situation lacks stability and those encountering barriers in access to home loans. Another differentiating 1 Some people consider the rise of the University population as an inflation of education levels, which do not generate gains overall, especially as during the education boom a lot of poor quality universities emerged that have granted Bachelor diplomas and are defunct by now. characteristic is that there is no need to cover initial costs. Buying a home is connected with the necessity to cover the down payment and often the costs of finishing or renovation as well as administrative procedures. In this respect, the other group who may be interested in renting will consist of people who are at the stage of gathering funds to fulfill this aim. Nevertheless it can be ascertained that tenants are usually individuals who can afford to buy their own house. In this sense renting does not fill the gap in satisfying the housing needs of people having varying incomes. This situation constitutes a significant challenge for housing policy which should start supporting the renting of private property at lower, subsidized rents. It must also be mentioned that unlike in many countries of the European Union, there is no cadastral tax in Poland. Tax burdens connected to housing are very low. In reality, therefore, citizens pay much less attention to the current maintenance costs of housing when considering buying a home.
In terms of urbanization in Poland, the unbalanced development of the rental market, as well as increased migration from rural to urban areas (both for university education as well as later settlement), mean that housing choices are rather limited with renting not being available to people living in rural and less urbanized areas while those who migrate to cities lose the option to live with their parents.
Our studies identify several practical implications concerning the shaping of social policy, housing policy as well as the development of new construction market offers. First, we believe that housing issues in Poland are not being solved and there are no solutions dedicated to people with lower incomes. Under current market conditions renting is not only less accessible to such people but also greatly increases the wealth gap between more affluent homeowners (since property prices in Poland have been growing significantly for many years despite amortization) and less wealthy tenants. In effect it is neither attractive nor does it bring any positive effects. This, in turn, inclines people with lower incomes to make risky decisions such as getting home loans that are at the border of their credit capabilities just to gain homeownership. In reality, the current system does not offer this group of people any solutions. There is also no efficient system for the monitoring of differences in the financial state, material situation, and the well-being of this group of loan takers which we perceive as a research gap and would like to pursue this direction in our analyses in the future. These types of risky borrowing decisions may result in a certain type of advancement in one's social and economic status since individuals do become homeowners, avoid the wealth gap, and are strongly motivated to remain active within the labor market. However, loss of financial liquidity caused by negative life events such as, for example, loss of work or the deterioration of one's health may have tragic consequences. In our opinion, there is a need to develop housing solutions for people in Poland who earn less which would protect them from such risks.
From the perspective of social policy, we would like to stress that, in our assumption, for as long as housing issues in Poland are not solved they will constitute a significant barrier to the reversal of the negative population growth seen in the country today. The implementation of programs meant to encourage Poles to have children should be supplemented with appropriate access to housing for groups of people with varying incomes.
When it comes to the investors' point of view within the market being considered it is necessary to point out that the housing needs of Poles are changing. The group of single people who would like to become independent by living on their own is growing. In our opinion demand for small flats that are well communicated with city centers and provide good access to employment will grow. It is also necessary for banks to prepare, both concerning risk calculation and administrative procedure, to attend to a growing group of loan takers who are either single or in informal relationships.
The question that emerges as a result of our analysis is whether the above changes can be compared to those that occurred in other post-transformation countries and those experienced by states of highly-developed Europe? Keeping in mind research presented in literature which shows that within two different countries and at varying times the occurrence of factors that may look similar but which are partially quite different is possible, we treat our article as an opportunity to invite researchers from other countries interested in this subject to join us in establishing international cooperation to address this. Considering that every country has unique economic and housing policy conditions, interpretation of results requires not only expert knowledge concerning econometrics and economics but, foremost, specialized knowledge of people who are very familiar with a given country.