32
2.7 An analysis of first-time buyers’ housing affordability by applying
33 In addition, in appendix 11 A, figure 16 , I demonstrate how the price- to- income index developed as a measure of housing affordability when defined in terms of the money households spend to buy reasonable housing (Arnold and Skaburskis, 1988). The fraction is based on real housing prices divided by real gross national income (and net disposable income) (for the data, see table 2 in appendix B). The increase in the index indicates the decline in housing affordability. The higher the index, the higher the proportion of income reserved to pay the housing price.
In 1992, the average house price level to aggregate income was higher by 350 per cent, a level, which was almost steady till 1998. However, at the housing peak years, the housing prices were higher than aggregate income by 700 per cent.
According to affordability development, I can see that more households will find housing purchasing less affordable. This should consequently have impacted on first time buyers’ buying decision and have limited housing demand. However, the demand and the prices continued to increase even further in 2006.
To conclude, the housing affordability development indicates that first-time buyers’ housing affordability declined significantly especially in 2003- 2006 (boom period). In 2007- 2009, housing affordability improved (bust period) due to a decline in housing prices, however, it still has not reached the trend we observed between 1992- 1998. This is a first indicator of housing affordability development- since 1998, the housing price growth does not coincide with gross national income or net disposable income growth.
The above analysis has provided valuable information on the evolution of the housing market.
However, it contains some weaknesses. Thus, the aggregate income, the denominator of the index, reflects the gross level income for the whole population in Denmark. It does not differentiate the income of buyers/owner-occupiers, who are a “target” group for measuring housing affordability.
The housing affordability for specific household types will be presented in the following.
2.7.2 Housing affordability across household types
In figures 4 and 5, I construct housing affordability across household types. Here, the price- income ratio is based on the median market price of a house/flat price divided by the median annual disposable and earned income for house/flat owner-occupiers (based on the data from “Investigation of Consumption”, see also tables 5 and 6 in appendix B). The analysis represents housing affordability for modelized families.
Also, the distinction between earned income and disposable income is interesting. The gap between earned income and disposable income is very important in the housing affordability concept. The
34 increase in households’ disposable income from one year to another indicates that there is a higher proportion of income to consume, or to pay for housing-related cost (for example, cost on a loan), leading to improved housing affordability, even if earned income is unchanged. Thus, the development in gap between earned and disposable income can also indicate how housing affordability has developed.
Figure 4 House owner-occupiers affordability Figure 5 Flat owner-occupiers affordability
Source: Statistics Denmark and own calculations
According to figures 4 and 5, since 1997, flat owner-occupiers must pay higher proportion of income compared to house owner occupiers, while prior to 1998, there was an opposite trend. This is explained by higher increase in flat prices compared to house prices (for another illustrative presentation see tables 17, 18 appendix 12A).
In addition, the analysis on households’ income for a corresponding dwelling type indicates that house owner-occupiers earn more (see figure 19 in appendix 12A). This is a second indicator of imbalances in housing affordability: since 1997, flat owner-occupiers pay higher proportion of income compared to house owner-occupies because of higher housing price and lower income.
(Here I disregard the sociological and psychological effects on increased demand for flats.)
In appendix 13 A, in table 2, figures 20, 21, I analyze more in depth housing affordability variables by growth ratios for flat and house owner-occupiers.
Generally, there was higher growth in housing prices compared to the growth in income. However, the developments, in general, are positively correlated.
Thus, the improvement in income disposable for flat owner-occupiers points to improved housing affordability in the median years 2001-2003 and after 2005 (can be also be seen in figure 5).
35 Therefore, it is a trigger for increased housing prices. The developments partly lead to an assumption that improved disposable income rather than earned income lead to increased housing demand.
In the years (1996-1999 and 2003-2004), housing price increase cannot be explained by higher growth in disposable income compared to earned income.
In bust times (2006-2008), a decline in earned and disposable income by 8, 93 and 26, 76 per cent correspondingly, reflected on declined housing affordability, as a result of income decline. Thus, the pressure on housing prices has already started in 2006. It might also explain the first decline in housing prices by 10, 78 per cent in the corresponding periods.
Thus, housing affordability for flat owner- occupiers and house owner- occupiers differs. Flat owner- occupiers pay higher proportion of income compared to house- owner occupiers.
Characteristics of location of dwellings have also an effect on affordability. How housing affordability developed across cities will be the topic for the next sub- chapter.
2.7.3 Housing affordability across regions
When I look at housing affordability across cities, house price development in the Copenhagen area, especially Frederiksberg, was characterized by higher increase in housing prices compared to other areas (see appendix 14A, figure 22). Here, of cause, limited supply of land and limited supply of housing affect the price equilibrium. However, taking income of corresponding areas into consideration, there is a big difference in housing affordability across cities (see figure 23 in appendix 14A). According to Shiller (2008) it is a “perceived scarcity of urban centers” effect on increased housing prices in urban areas. But, according to Shiller (2008), it is a myth that scarcity of urban/ metropolitan areas and unmet demand for it should boost housing prices in metropolitan areas. This demand can in fact be met- new urban centers, new towns or cities can be built from scratch when there is a demand. It requires only sustained and coordinated effort (Shiller, 2008).
Therefore, the prices for a house in a big city should not be different from a price of a house in a smaller town. However, he assumes a supply side with high and swift elasticity, which is not a reality (or, at least, there is very limited possibility for doing it).
In addition to this, according to Ball (2007), it is not clear from a rational perspective, why house prices should rise in such places as Amsterdam, London and Madrid. I would like to add Copenhagen. Thus, the significant increase in Copenhagen/ Frederiksberg cannot be explained by increase in earnings. This is a third indicator of imbalances in housing affordability development- the owner- occupiers in big cities pay higher proportion of income on housing compared to smaller cities’ owner- occupiers.
36 (Here I disregard the increased demand for housing in metropolitan areas because of sociological and psychological preferences, location to jobs, schools and day- care, as well as the lower supply of housing in capital centers. Only income level is assumed to affect the demand.)
Thus, the analysis of the evolution of housing prices by applying the housing affordability approach has indicated that the general level of housing prices was overvalued. Especially flat owner- occupiers have to pay higher proportion of income compared to house owner-occupiers. Also, big cities’ owner-occupiers have poorer housing affordability compare to smaller towns’ owner- occupiers. However, the analysis of affordability has some disadvantages.
2.7.4 Critique on price-to- income ratio as a measure of housing affordability
1) The price-to-income ratio disregards other variables that affect housing affordability. According to Yang and Shen (2008), housing affordability is: “a subjective and complex concept than cannot be neatly or simply assessed by a single ratio of house price to income”. The concept reflects the households’ decision process: “it is a function of decisions that households choose to make between housing expenditures and non-housing goods” (Yang and Shen, 2008, p.321).
In this view, “there are three critical dimensions of housing affordability: income, non-housing demand and housing demand” (Yang and Shen, 2008, p.318). Therefore, the cost of non-housing necessities, housing-related costs, financing costs should also be considered when measuring households’ ability to buy a house (it will be the subject for chapter 2.8).
2) The average housing affordability index of value of 100 has been derived on the basis of long- term price-to-income ratio for the period 1992-2008. An average ratio, based on a longer time horizon may lead to different results.
3) The analysis on the aggregate level can lead to misleading results, because affordability problems among different households’ demographics may have different consequences for housing market and financial stability. For the analysis on housing affordability across demographics, see appendix 18A.
4) The housing demand can be driven not only by the current and expected housing price level, but also by the past housing price development. This makes dubious the assumption that improved/declined housing affordability increases/decreases housing demand. I will discuss it in the following sub-chapter.
37 2.7.5 Reflections on the basis of behavioral finance assumptions
The demand for credit/housing is conditional upon perceptions of the current and future values of housing and of other goods (Miles, 1994). However, past housing price behaviour also influences the demand for credit and house. Case and Shiller (1989) show for instance that extrapolating behavior (backward looking expectations) is common in housing markets. During booms home buyers expect further housing prices rise and are worried about not being able to buy a house in the future market (Shiller, 2007). Accordingly, the backward-looking expectation influenced affordability assessment.
The developments in housing prices during boom times influenced peoples’ perceptions and expectations. There were beliefs that housing prices will always be in surge, which in turn increased borrowing capacity. According to Akerlof and Shiller (2009), there were no rational arguments about investment possibilities. The only argument was that investment in real estate was a spectacular investment because prices will only go up. This strong intuitive and naive feeling was among people all over the world. The spread of this argument by mouth-to-mouth fed the boom during decades (Shiller, 2008).
The “money illusion” appeared to explain some of the impression that homes are spectacular investment. “People compare home prices from the purchase time, however, they tend to forget to compare with other goods or even forget about inflation. However, the real value of the home may have only doubled over that interval, which would mean an annual appreciation of only about 1.5%
a year” (Shiller, 2008).
So, as prices go up, so does investors’ confidence. According to Kindleberger and Aliber (2005), this confidence (the euphoric behavior) sparks asset prices even further because the eagerness to buy is stronger than the eagerness to sell.
It can therefore explain the fact that declined affordability (as a result of higher house price) sparks housing demand even further- people look at the historical house price increase and form the belief in constant house price increase. Thus, they rush to buy a house now because it might be less affordable in the future. This behaviour explains why housing prices are not correlated with the current income level and contradict my main assumption that changes in housing affordability should affect housing demand.
So far, the analysis of housing affordability was based on two variables, the income and housing price level. In the following section, I discuss how housing affordability should be defined not only in terms of income, but also in terms of all housing-related cost for existing home-owners. The analysis derives the historical level of housing-related costs. It is necessary to include these into housing affordability assessment because it demonstrates households’ ability to sustain a house.
38