Wealth Taxes and Private Residential Real Estate

The chatter on wealth taxes has been increasing in intensity over the past year. The reason propounded for an implementation is to promote of an inclusive society by finding additional sources of government revenue to fund various altruistic schemes. The shift away from taxing income to that of wealth is perhaps a way, but not the only means to achieve that objective. From public forums, the form that the wealth taxes being considered here appears to be focused on the taxation of property gains and/or inheritance.

In this blog, we shall examine whether a wealth tax on private residential property is ideal, discuss some alternatives and if such a tax is to be levied on residential property, a suggested form it should take. The analysis will look back to 20 years in time to the periods 1H/2001, 1H/2011 and 1H/2021. (We shall not be discussing on the forms of wealth taxes nor how it should be implemented. We will also not venture to areas outside of the taxation of private residential real estate.)

Firstly, we studied how the Lorenz curve for private residential transaction values has evolved over time. Graphs 1a, 1b and 1c shows the Lorenz curves for the three periods. From this, we can calculate the Gini coefficient for the transaction market. (The Gini coefficient is the area of the portion bounded by the blue and orange lines in the graph divided by the total area under the orange line and the x-axis.) The interpretation of this Gini coefficient is slightly different from that used to describe household income. It is: ‘The seller of the property receives cash in the form of equity and/or profit (if any) from the transaction. The larger the Gini coefficient, the bigger the disparity in sales proceeds (and from the perspective of the buyer, his/her buying power) in the transaction value curve vis-à-vis a uniformly priced private housing market.

Optically, the Lorenz curves for the three periods do not appear to have varied too much. That is in fact almost true. Table 1 shows the Gini coefficients for the private residential transaction value market and national household income. For private residential transaction values, it had ranged between 0.255 and 0.304, lower than the 0.375 to 0.433 range for household income for these three periods in the last twenty years. The ratio of the two shows that private residential transaction value inequality varied from 0.61 to 0.77 for these three periods. As transaction values map onto the valuation radar which in turn gets pushed onto the private residential value topology, it hints that our private house values are closer to the ideal uniformity compared to income levels.

If we break down the private transactions data to show the New Sale and Resale transactions, Graphs 2a to 2c show that the Lorenz curve for the resale market in H1/2021 had strayed more from the line of equality. Table 2 shows this in terms of the Gini coefficients.

If this is the trend moving forward, this raises the question of whether a common rated wealth tax levied across the private residential market would be ideal or it should be on the top end of the value transacted to minimize the value distortion.

Now, we turn to the HDB Resale market and observe how the Lorenz curves have been shaped over the three periods in time. Please refer to Graphs 3a to 3c and Table 3 for the Gini coefficients. The measure on equality has been improving over time although in H1/2021, it had deteriorated over H1/2011. However, the Gini coefficients for the HDB Resale market are lower than those in the Private Residential market.

For a better comparison of the housing market to income, we combined both the HDB and Private Residential transaction data to generate another Lorenz curve. These are shown in Graphs 4a to 4c with Table 4 showing the Gini coefficients.

Table 4 shows the combined HDB Resale and Private Residential Transactions Gini coefficient for the three periods and how they quotient against the income counterpart. The two are quite similar and although the housing coefficient has increased over the decades, it is quite congruent with income. There are those who may ask why the Gini coefficient for the combined public and private markets is greater than its components, it is due to what is called Simpson’s Paradox.

Furthermore, if we look at the absolute transacted prices for HDB and private residential, there is considerable overlap. Graph 5 shows that in H1/2021, 39.3% of private transaction prices intersected with HDB resale prices. This was an improvement from the 31.3% in H1/2011 although it is still about 10 percentage points lower than what it was in H1/2001. The slate of cooling measures implemented in the past ten years may have prevented prices in the private end of the housing market from running away, causing the overlap to increase again.

Looking at the average household income levels for those who lived in non-landed and landed private residential homes, we find that the upper income groups have seen their numbers increase at the three periods. Graphs 5a to 5c show that the highest income sub-group has increased. The graphs have been plotted for the minimum average income and above for households who live in private residential properties. The minimum average income for the three periods is shown in table 5 and given that landed property dwellers have a lower income (surprise!), that will be used as the starting point to plot Graphs 6a to 6c.

The sharp increase in the number of households in the highest income bracket could be for several reasons. Some of which can be the importation of senior managerial professionals, both foreign and local entrepreneurs who made their fortunes, higher number of locals moving up the income ladder etc. (Please refer to Graphs 7a to 7c.)

However, the swelling in the highest income subgroup numbers has not elastically translated into rising private residential property prices. Table 6 shows that private residential prices have been increasing at a very modest compounded growth rate for the past 10 years. For non-landed prices in the luxury segment, they have instead fallen. From 2010 to 2020, median household income (a CAGR of 3.8%) and nominal GDP (a CAGR of 3.7%) growth have outpaced private property price increases (a CAGR of 1.2%). (Please refer to Graph 8.) In fact, for the top 10 listed developers who have been active in the local residential market in the past 10 years, their shareholders funds, a proxy of their financial strength, has been rising at a CAGR of 6.2%. (Could this be the reason why land prices have been bid up aggressively but, because of the cooling measures, developers could not be aggressive on their selling prices, leading to falling profit margins in the last 10 years?)

We now turn to those in the top income bracket (ultra-net worth individuals). Using the land price for Good Class Bungalows (GCB) as a proxy for their housing choice, we find that it is negatively correlated to the national income Gini coefficient. (Please refer to Graph 9.) While there may be many ways to interpret this, including one that simply says that this is a spurious correlation, we believe that it does give some hint that the ultra-high net worth class here has generally been contributing to the welfare of society while not being penalized (as reflected by the value of their GCBs) for doing that. (A negative correlation is good because it means that although the rich are reaping capital gains (GCB land prices rose at a CAGR of 3.8% from 2010 to 2020), income disparity in general has also been narrowing.)

The slower price increase for island wide private properties and a negative growth rate for the luxury segment is likely attributed to the slew of cooling measures that were implemented in the last decade. This then begs the question of whether a wealth tax on real estate is justified? For this matter, we list down some of our thoughts:

1. We believe that there is already a form of wealth tax levied on private residential properties. That comes in the form of the Additional Buyers Stamp Duty (ABSD) that is levied on those who can afford to buy more than one home. There is also a form of capital gains tax that is labelled as the Sellers Stamp Duty (SSD).

2. Linked to the first point, if a wealth tax is imposed on private residential real estate, then the ABSD and/or the SSD may have to be recalibrated to offset the tax. The reason? Imposing a wealth tax on top of the two would unduly yoke the market and may inflect it to a dis- or deflationary state in the long run.

3. If a wealth tax is imposed on the highest valued private residential properties e.g. GCBs, the question is how much would the overall amount be? In 2020, about S$2.5 billion worth of GCBs were transacted (source: Realis). As a comparison, the nominal GDP for the Services industry in 2020 amounted to S$313.7 billion. A tax on the top level of the housing food chain would not add much to government revenue.

4. As our population ages, we also find more of them are living in private residential properties. Graph 10 shows the percentage of residents aged 50 and above living in non-landed and landed private residential properties. Given the subdued private home price increase in the last 10 years, due consideration should be made as to whether a wealth tax on private residential properties may reduce the cash back to the aged should they wish to downgrade to cheaper housing to free up cash for consumption or medical expenses.

5. Technology is a two-edged sword. While it brings about untold benefits, it can also easily lead to many jobs being made redundant and skillsets irrelevant. Those who are presently paying the mortgage for their private residential properties may for now able to do so because they are gainfully employed. However, we are now living in a Volatile, Uncertain, Chaotic and Ambiguous (VUCA) world which can invert one’s career path without even knowing it until it is too late. Aside from functioning as a roof over one’s head, the home doubles up as an investment that one can draw down on when life’s fortunes are vectored in the wrong direction. There is only so much re-skilling can do because although different types of intelligence peak at different ages, those associated with the sciences tends to do so early (Source: Scientific American – When Does Intelligence Peak by Scott Barry Kaufman on Feb 28, 2019). Therefore, in an age where technology increasingly takes to the economic and social centerstage, structural unemployment is likely to come earlier for many. That dreaded housing downgrade may be necessary. The equity and profit released in such a move could be a lifesaver for some and therefore ideally, should not be taxed.

Conclusion

Intuitively, wealth taxes on private residential properties looks like a low hanging fruit that can add additional sources to government revenue. It is attractive because prima fascia, it appears to achieve the objective of transferring wealth from the upper spectrum of society to help those at the lower end. However, looking at the private housing market, from the way values are spread out through the value chain to how prices have performed over the past decade, the structural changes that our middle to upper middle-income earners are increasingly being confronted with, and the lowering of the intelligence quotient due to age, the idea of imposing a tax in a direct form (capital gains, inheritance tax on residential) needs to be much more seriously discussed.

Amongst others, one issue with having another tax on residential property here is that there are already taxes and stamp duties. The latter was imposed to cool the market, subduing price increases over the past ten years. But if we relabelled these measures, they would also function as a form of wealth tax on private residential real estate. A further tax, without recalibrating the existing duties, may tip the market to a disinflationary or deflationary state.

The need to render assistance to the lower income and mentally challenged groups has been a perennial problem. But the road ahead will be more volatile, unambiguous, complex and chaotic. Analysts and economists have often cited one downside of technology as accelerating job redundancies. In reality, the situation is even more dire. The path ahead is not linear nor smooth. Along the way, there are economic crises, and now, on our doorstep, climate change. Each time we go through such interludes, people become economically displaced, and most may not be able to loop back to the size of their last paycheck drawn. It could very likely be the middle to upper middle-income classes of today that ultimately become those that need help. And because of their legacy household expenditure patterns, are not easy classes to assist. For them, whose careers have been torpedoed by technology or crisis, their best help is a housing downgrade to thaw out their fixed assets. However, because a larger proportion within these classes also live in private homes, a wealth tax on their property would reduce the amount of self-help they would get.

In a VUCA world, the need for sources of funding to help society is becoming even more pressing. Many schemes come with Catch-22 situations and we do not pretend to have a compendium of solutions. Sticking to the private residential market, we know that it has already been overlaid with many measures meant to prevent a price overspeed. With private home prices rising at a 10-year CAGR of 1.2%, the market has very few degrees of freedom to move higher. Wealth taxes would further reduce the market’s latitude on the upside and risks tipping it down instead. But home prices need to rise because buying a home is an investment. Nevertheless, if the finality is such that a wealth tax must be implemented, then some of the slew of cooling measures may have to be rolled back to give the market slack to return to an even keel.