3.3 Historical and Reference Data
3.3.2 Plot Ratio
Plot ratio, the ratio of building gross floor area to site area, is one of the most important factors affecting land use footprint, and because it is directly set by zoning policy it is one of the strongest levers the Singapore government has to influence land use efficiency. This section details how plot ratio has been used to decrease residential land use footprint throughout the nation’s brief history.
A survey of Singapore’s housing stock identified 12 typologies of housing, which were physically modeled to study and compare built forms, and quantitatively an- alyzed for their plot ratio, site coverage, and height in stories. Photographs of the physical models are shown in perspective in Figure 3-6, identified in plan in Figure 3-7, and compared quantitatively in Figure 3-8, using the SpaceMate chart [67].
Before 1960, Singapore’s housing stock consisted of two primary building typolo- gies: the kampong and the shophouse. Kampong houses are a traditional vernacular structure often raised above ground on pilings, of timber construction and a thatched attap or corrugated metal roof. Units are typically clustered together to form infor- mal streets and public spaces, see Figures 3-6 and 3-7, first row, left. The shophouse
Figure 3-6: Typical Singapore housing typologies: form
is another traditional southeast asian housing prototype, typically a two story ma- sonry construction with shop space on the ground floor and living space on the upper floor. Units are arranged densely, in block-long rows with shared party walls, covering nearly all of the available site area, see Figures 3-6 and 3-7, second row, left. While a few neighborhoods of these housing typologies remain today, neither model was suffi- cient to house the ballooning population of post-independence Singapore. In fact, to alleviate the dangerously crowded conditions found in many such neighborhoods the URA carried out an urban renewal program to demolish and redevelop most of these areas, meaning the traditional housing typologies play a negligible role in residential land use today.
Figure 3-8: Typical Singapore housing typologies in SpaceMate chart, comparing site coverage, plot ratio, and height
Upon gaining independence, the Singapore government turned its attention to the rapidly increasing population and severely lacking housing supply. Rather than allow private development to propagate the existing kampong and shophouse models across the entire island, which would have limited plot ratios to less than 2.0, the government pooled its resources to form the Housing Development Board, an entity with sufficient funding and power to establish a comprehensive public housing program. Referencing Leung’s [68] study of the HDB’s design evolution, between 1960 and today there have
been five distinct generations of HDB housing typologies, defined in Table 3.3.
Table 3.3: HDB generations, adapted from [68]
Gen. Description Time Plot Ratio
I Composition of simple generic blocks 1960–1966 2.0–2.8 II Building blocks as a tool for urban design 1967–1981 2.4–2.8
III Tightening of sizes of precincts 1982–1990 2.8–3.0
IV Diversification of design 1991–present 2.8–6.0
V Super-high density redevelopments 2000–present 3.5–12.0
While the extremes of HDB plot ratios vary greatly over the five generations, Leung states that the average plot ratio has remained very near 2.8 between 1960 and 1990. Additionally, a qualitative review of the Master Plan 2008 [69] shows most public housing sites zoned for a maximum plot ratio of 2.8. Since 1990, however, more HDB projects are being built at higher plot ratios, especially redevelopments near the city center. In the model, it is therefore assumed that the average plot ratio of the existing stock in 1990 is 2.8, but the input plot ratio increases linearly from 2.8 to 3.0 between 1990 and 2011, owing to an increasing fraction of new HDB buildings being of the higher density redevelopment type. See Tables 3.4 and 3.5 for the initial and input plot ratios, respectively.
In order to retain some variety in housing stock, as well as to appeal to the affluent and upwardly mobile population, Singapore has always maintained a small share of high-end private housing. Private housing comes in two distinct varieties: landed and non-landed. Landed housing is by far the most expensive and least dense housing class in Singapore. Even with its exorbitant pricing, a study of residential land utilization criticizes that the government is losing valuable housing potential and revenue through land sales for such low density development [70]. As a result, the total amount of landed private housing has been strictly limited, and is unlikely to grow significantly in the future.
Before 1990 private housing represented a mere sliver of the total housing stock, but since 1990 the provision of new non-landed private housing has been consistently
increasing, even while the provision of new public housing units fell sharply after 1999, see Figure 3-3. The result is that for the past decade new private housing units have exceeded new public housing units in sheer numbers. This means that the private fraction of the total housing stock is growing, significantly. So today, more than ever, the density trends in private housing have a profound effect on total residential land use.
Data for private housing plot ratio was gathered from government land sales for residential development [71, 72, 73]. The data show no particular trend in density change for landed properties or apartments, so the average of all recorded plot ratios was taken for each property type and recorded in Table 3.5. Also, since no trend was identified, it is assumed that the average plot ratio of the housing stock in 1990 was equal to the current value, see Table 3.4. There is, however, a distinct trend in increasing density of private condo developments, shown in figure 3-9. New condo- miniums increase linearly in average plot ratio from roughly 2.2 in 1990 to about 3.5 in 2012. This change serves to decrease the residential land use footprint over the modeled period.