Differing conclusions on where value of flats is likely to be lower
By Jessica Cheam
ARE the values of public housing flats in opposition wards lower than in those of the ruling People's Action Party?
This was a contentious issue that surfaced during the recent General Election. Everyone from party leaders to members of the public took different positions, using statistics to back up their arguments.
Former minister mentor Lee Kuan Yew started the debate when he said that homes in Hougang, held by the Workers' Party (WP) since 1991, sold for less than similar ones in neighbouring areas because they belong to an opposition ward.
WP chief Low Thia Khiang rebutted these comments by citing figures from the Housing Board website that showed five-room flats in Hougang SMC sold for higher prices than those also in Hougang but located in the PAP-held Pasir Ris-Punggol GRC.
In response, former national development minister Mah Bow Tan pointed out that Mr Low had used flats in Pasir-Ris Punggol GRC for comparison, when the constituency nearest to Hougang SMC is actually Aljunied GRC.
Mr Mah then highlighted figures from the HDB website showing that five-room flats in Aljunied GRC fetched 15 per cent more than those in Hougang over the past 12 months.
Four-room flats in Aljunied GRC had a 16 per cent premium over those in Hougang, while three-room flats were sold for 3 per cent more.
The units in the calculations were located in Hougang Avenue 6, 8 and 10 for those in Aljunied GRC, and in Hougang Avenue 5 and 7 for those in Hougang SMC. The blocks were on either side of Upper Serangoon Road and roughly equidistant from the MRT station. They were also the same age - having been completed between 1980 and 1985.
In its analysis, the Ministry of National Development (MND) also pointed out that within Hougang Town, flats in Hougang SMC generally have a locational advantage over those in Ang Mo Kio GRC and Pasir Ris-Punggol GRC, as they are nearer the town centre and the MRT station.
'This is why it is more meaningful to compare flats in Hougang SMC with flats in Aljunied GRC, and even then, only between those in comparable vicinities,' it said, justifying why it limited its study to the blocks it chose.
THE study prompted other analyses that strove for the opposite conclusion.
One notable study was by two PhD students and trained statisticians, Mr Chong Kwek Yan and Mr Giam Xingli, from the National University of Singapore and Princeton University respectively. But both made clear the study was their own, not that of their university.
The duo found that if one looked at 15- to 20-year-old flats, HDB's website data showed that flats in Hougang SMC were nearly 4 per cent higher in value than the same type of flats in Aljunied GRC. HDB's website makes available resale transactions of flats up to a year ago.
They argued that Mr Mah's analysis, which uses only 25- to 30-year-old flats, was not a meaningful comparison as looking at a specific group of flats by flat age is too narrow a comparison.
To capture the widest possible data set, the researchers broadened their calculations to include all transacted five-room flats of all ages from Hougang SMC and flats in the Hougang part of Aljunied GRC.
They also used statistical techniques to control for age of flat, size of flat, type of flat, and height/level of flat, all of which they said are 'factors that co-vary strongly with the value of a flat'.
Their conclusions: A resale flat in Aljunied-Hougang can be sold, on average, for between $22,600 more to $3,600 less than a resale flat in Hougang SMC.
Using the same methodology, however, a resale flat in the Hougang Town portion of Ang Mo Kio GRC was worth on average $26,500 less than a flat in Hougang SMC.
Similarly, a resale flat in the Hougang Town part of Pasir Ris-Punggol GRC was worth $18,000 less than a flat in Hougang SMC.
So who's right?
BOTH arguments have their merits, say statisticians.
'They're both right. It's normal for researchers looking at the same data to draw different conclusions,' said Associate Professor Low Chan Kee, acting head of the economics division, School of Humanities and Social Sciences, Nanyang Technological University.
'There are so many factors in determining the value of flats and these assumptions will affect how researchers conclude their studies. The important thing is for readers to judge for themselves, to decide which of the two studies reflect what they look for in a flat - different facilities, the age of the flat, the location,' added Prof Low, who specialises in applied statistics.
While Mr Mah's analysis controlled for age and type of flat, as well as location, one could argue that the data set was too small to draw definitive conclusions.
The two researchers' study, on the other hand, included a much larger set of flat transactions, including all flats aged about seven to 36 years old and comparing across four different constituencies.
But its analysis had one major flaw in that it did not account for locational differences.
When contacted, Mr Giam said they were unable to test if locational differences - distance from amenities and transport nodes as defined by HDB - account for price differences as geographic information system (GIS) data of flats is not publicly available.
'If the HDB is willing to release the data, we are willing to incorporate location in our analysis,' he told The Straits Times.
But he noted that if the HDB wanted a more comprehensive study, it could have compared prices of flats in Hougang SMC with those in Aljunied, Pasir Ris-Punggol and Ang Mo Kio GRCs that are the same distance from the nearest MRT station.
'Simply comparing Aljunied GRC and Hougang SMC also does not tell us any general relationship about political affiliation and flat prices, because no matter how many flats they include, it's still effectively a sample size of two,' he said.
If the same relationship is found between Hougang SMC and its neighbouring PAP-held GRCs, as well as Potong Pasir SMC and Bishan-Toa Payoh GRC, then perhaps one can conclude with more confidence that the political party controlling it has an effect over and above all other factors, he added.
'But if the trend that they claim only holds when comparing Aljunied and Hougang SMC and not between other pairs of opposition versus PAP comparisons, then it may be something due purely to other unique differences between Aljunied and Hougang.'
Prof Low said both studies needed a wider scope to lead to appropriate conclusions for a statistician.
'If they want to draw a fair conclusion they need to conduct a very thorough study involving a lot of different factors,' he said. 'The study also needs to take into account whether the difference in flat value is significant and the possibility of sampling error.'
Mr Giam threw up a different suggestion: The best way to assess the impact of opposition control on flat prices is to monitor the flat prices of Aljunied GRC over the next five years or so.
The Workers' Party clinched a historic win when it won Aljunied GRC in the recent election - the first time an opposition party has wrested control of a GRC from the PAP.
'We all know that flat prices are likely to increase (with the general economy), but will the increase of flat prices in Aljunied and Hougang SMC keep up with the increases in neighbouring PAP constituencies like Ang Mo Kio GRC and Pasir Ris-Punggol GRC,' asked Mr Giam.
'Knowing this may help to advance our current understanding on how political parties affect flat prices,' he said.
But until more such data sets become available, the debate is likely to continue.
Additional reporting by Cheryl Ong