Factors influencing housing price differences in the greater Athens area
During the last few years, real estate prices have increased at rapid rates in many different countries, including Greece. Theories suggest that housing prices are related to interest rates, rents, construction costs and other determinants of supply and demand, such as location characteristics or the attributes of the population living in a region, like income or wealth. This article tries to uncover some of the factors responsible for the differences in housing prices across the different districts of the Athens metropolitan area. It focuses on the variables which vary across the districts in a specific time period. The causes for such differences must be found in characteristics that change across the regions in any given time period. I find that the primary determinant for the price differences is the differences in rents, but that the price-to-rent ratio varies across the districts according to the potential housing supply, the level of commercial development, as well as population characteristics, such as income, wealth, social status or ethnicity. This article combines two different sets of data. The first set is by Professor Epameinondas Panas, of the Athens University of Economics and Business and is available on the Internet (www.ared.gr). Professor Panas splits Attica into approximately 40 districts and reports consistently the median price and the median rent per square meter in each area. The second set of data comes from EFG Eurobank, which conducted a survey of Greek households from December 2005 – January 2006. The survey contained questions regarding income, wealth, education level, households’ predictions of the future course of prices and the characteristics they value most in a house. Relation of prices to rents Diagram 1 reveals a strong positive association between housing prices and rent. It contains a cloud of 40 data points, each point representing the combination of prices and rents of any given region. The cloud has a clear positive tilt, with a slope of 32.7: Districts with, say, a higher median annual rent per square meter of -1, are associated with a higher median price of -32.7 per square meter. In percentage terms, if rents in one district are higher than in another by 10 percent, prices would then be higher by 11.3 percent. Each data point in the cloud deviates substantially from the positive trend line. These deviations suggest that besides current rents, there are other economic forces influencing the level of prices, as follows: Supply seems to play a role in the price-to-rent (PR) ratio. Survey respondents were asked to rank their willingness to sell and it turns out that the willingness to sell is higher in districts of higher prices relative to rents. On the other hand, households seemed to be more willing to move to a bigger house in districts with relatively lower PR ratios, suggesting that these are the areas for potential growth. Wealth is an important factor behind demand. Survey respondents were asked to fill in the category of wealth they belonged to from 1 to 8, 1 being the poorest (less than -30,000 a year) and 8 meaning the wealthiest (more than -1.75 million). Diagram 2 shows that regions with wealthier households are also areas with higher PR ratios. A step up in category is associated with an increase in the PR ratio by 1.68. A similar relation holds for disposable income. Districts with households in the higher income brackets are regions with higher PR ratios. When it comes to real estate, the popular phrase is: «Location, location, location.» Households were asked to rank their preferences from 1 (not important) to 5 (very important) on a number of location characteristics. Surprisingly, most location characteristics do not seem to be related to the PR ratio: the density of traffic, the crime rate, proximity of the house to schools, the center of town, work, transportation, parks and/or athletic facilities. One interpretation of the absence of a relation is that in the greater metropolitan area of Athens, many of these factors do not differ enough from district to district to interest homeowners. The analysis did identify one location factor as being important to the PR ratio: The higher the households’ desire for low commercial development, the higher the PR ratio. Social status and country of origin seem to also be related to the PR ratio. Social status is negatively related to the PR ratio, suggesting that people who belong to higher social classes tend to drive a hard bargain when it comes to prices. More interesting, however, is the relation between PR ratios and the country of origin of the head of the household, i.e. whether (s)he is Greek or an immigrant. About 10 percent of the sample respondents were immigrants, but the percentage varies from region to region. It turns out that regions with more immigrants are regions with higher PR ratios, suggesting that immigrants tend to be buyers in greater proportions than ethnic Greeks, driving up the prices relative to rents in the districts in which they tend to be more concentrated. Conclusions The above analysis shows that in the Athens metropolitan area, the level of housing prices is positively related to the level of rents. Other factors positively influencing housing prices are the wealth and income of the households living in any given district, the low commercial development of the district and the proportion of immigrants living in it. Surprisingly, factors such as traffic, proximity to schools and work, or the crime rate were not found to be important. (1) Angelos Hardouvelis is a candidate for the International Baccalaureate Diploma.
