Mr Speaker, in my speech, I wish to explain why raising GST is a bad idea at this point in time, how it will also be bad for working families, and what we can do as alternatives to raising GST, to meet our fiscal needs.
The economic environment is turning less favorable
Minister Wong began his speech by painting a cautiously optimistic picture of both Singapore’s as well as the global economy. We grew by 7.6 percent last year, exceeding even the upper bound of the MTI’s forecasts.[1] Unemployment also plunged, to 3.2 percent, close to pre-COVID levels. And exports—the traditional engine of growth for our economy—expanded by around 12 percent. With growth, employment, and trade figures strong, there is some justification for a renewed sense of confidence as we emerge from the pandemic.
But the economic picture is not all peaches and roses. In January, inflation clocked in at 4 percent compared to prices a year ago, and the highest rate in almost a decade. Higher prices have been exacerbated by an overall increase in the cost of living in the city, which has already been rated as among the most expensive in the world.[2] And tailwinds from the generous fiscal stimulus as well as a favorable external environment will fade further, with growth, accordingly, expected to fall to around half that of last year.[3]
Now is a bad time to raise GST
It therefore strikes me as premature to be contemplating any form of tightening in the stance of fiscal policy at this time. While the removal of wage support for all but the most hard-hit sectors of the economy is both prudent and economically sound, going in the other direction—of making fiscal policy more contractionary—takes a sensible argument to an absurd conclusion. Doing so would introduce unnecessary frictions into what is still a nascent recovery. It runs the risk of shooting ourselves in the foot, scoring an own goal.
But this is precisely what will occur with an increase in our goods and services tax (GST).
The example of Japan is prescient here. Japan applied consumption tax hikes—their equivalent of GST—three times in recent history, first in 1997, then in 2014, and subsequently in 2019.[4] The first two times, GDP collapsed, and the economy promptly went into a recession. Inflation also jumped. What’s worse, the subsequent collapse in output meant that the tax increase itself failed to produce sufficient revenue to justify its increase.[5] The most recent time was also accompanied by a recession, although this time, the effects were exacerbated by the onset of COVID-19. While the most recent Omicron variant has turned out to be less detrimental to economic performance than earlier ones, this episode is a reminder that a GST hike would rob us of the precious opportunity to revive support for the economy, if—knock on wood—some unforeseen pi or rho variant (I’m showing off my very limited knowledge of the Greek alphabet here) were to emerge.
In Singapore, we last hiked GST in 2007. This did not trigger a recession the following year—the economy eked out a 1.1 percent growth rate in 2008—but it nevertheless represented a sharp drop in activity, compared to 7.8 percent the year before.[6] Back then, we were helped by the fact that in both years, global growth remained in positive territory.[7]
This time round, the global environment today may be less favorable. The International Monetary Fund forecasts a slowing of world GDP growth—from an anticipated 4.4 percent this year—to 2.0 percent in 2023, and 1.6 percent in 2024.[8] Thus, global growth will be slowing, just as the burden of increased GST will be kicking in.
A hike in GST is unarguably a contractionary fiscal policy that could offset what has hitherto been very healthy consumption and investment growth over 2021.[9] The private sector—by anticipating the impending tax changes, and the need to save more—could further scale back on spending today.[10],[11]
A GST hike is also particularly ill-timed, because it holds the potential to stoke already-elevated inflation. While the direction of inflation in the future is anybody’s guess—economists have a particularly dismal track record of predicting price dynamics, one that would make a weather forecaster blush—it would not be surprising if inflation remains elevated through year-end. Supply chain constraints have not resolved as quickly as previously anticipated. Energy prices are high, and could rise more in response to geopolitical tensions and the likelihood of global sanctions. Labor markets are tightening, spurred by global Great Resignation and 躺平 (or “lying flat”) movements. Indeed, MAS expects as much, having raised its latest forecast to the 2.5–3.5 percent range.[12]
Moreover, this state of affairs may even become a more lasting fixture. There is a risk that inflationary expectations have become unmoored from the traditional 2 percent target held by most advanced economies worldwide, and is now permanently higher by up to half a percentage point, compared to before the pandemic.[13]
If this turns out to be the new reality, then rolling out a GST hike at this time will only add fuel to the inflation flame, which is already starting to burn uncomfortably hot.
It is for this reason that any delay of the hike—in contrast to the historical pattern of rolling out increases by July[14]—is critical. Doing so in a staggered fashion—1 percent in 2023, and another in 2024—will also cushion the impact. This measured approach is broadly supported by local economists and observers, and SMEs have also expressed their relief about the delay.[15]
The key question that remains here is this: Will the government stand ready to potentially reexamine the timing, with a view toward a possible delay, should macroeconomic conditions remain unfavorable for a hike come 2023 and 2024? And should the government proceed with a hike at the preannounced dates, will they complement the hike with a justification based on the contemporaneous economic situation?
It will be bad for the worker if GST is raised
But why raise GST at all?
The reason, provided by the Ministry of Finance, is that our revenue needs have risen, especially with regard to providing for healthcare in a rapidly-aging population. There is no doubt that we should do more to take care of the elderly among us, the Merdeka and Pioneer generations that have contributed so much to bringing Singapore to where it is today, as well as generations that will enter their twilight years in the future.
My concern, Mr Speaker, is whether the GST is the ideal instrument for doing so.
On its face, GST fulfills many of the criteria economists typically look for in optimal public finance. It is efficient, because taxing all consumption equally does not distort spending behavior compared to taxing income or other taxes specific to certain goods or services.[16] It is broad-based, meaning that it can raise significant revenue across a large population.[17] And it is simple: unlike income, we do not need to futz around with appropriate measurement and exemptions; you pay a fixed fraction of what you buy, period.
This is precisely why we do not rely on the GST alone to raise income. It is generally sound to diversify the government’s revenue stream, as opposed to relying on narrow forms of taxation. More recent research has questioned whether consumption taxes are truly more efficient, especially if we relax some of the standard assumptions.[18] But perhaps most importantly, most people believe that it is easier to ensure that income taxes are progressive: that is, those who are better able to afford a higher tax amount pay more. The GST is a flat tax rate applied to all purchases of goods and services, which hits lower-income households disproportionately more. In and of itself, it is therefore regressive.
The government’s solution to addressing regressivity has been to pair GST with a voucher system, where cash rebates of a certain amount—albeit not matched directly to actual spending—are issued to lower-income households. This transfer is indeed progressive, and it may be the case that the overall tax system—after raising GST—remains progressive. But an increase in GST will, in and of itself, reduce progressivity, which means that we end up less progressive than the status quo. I therefore welcome the Minister to clarify if the plan to raise GST while expanding the scope of the permanent GST voucher (GSTV) scheme would, on net for the nation as a whole, be more or less progressive, relative to not increasing the GST at all.
This is because we all understand that the generous $6.6 billion Assurance Package—while certainly welcome during the planned 5-year transitional period—will not last forever. After all, the earnings of lower income groups have been eroded comparatively more as a result of the pandemic, and furthermore, and support payments via the COVID-19 recovery grant will expire by year-end.
I have focused so much of my remarks on the importance of supporting the less fortunate, not just because of some heightened notion of equity. Many studies have pointed to how pandemic-induced changes will have lasting effects on the incomes, savings, educational attainment, and other socioeconomic outcomes of poorer households.[19]
Mr Speaker, the poor were already relatively behind prior to the pandemic. As a society, we can ill-afford to have the fall even further behind. It will be an affront to our sense of equality of opportunity and genuine meritocracy, as well as a stain on the morals and values we hold as a society.
Finally, it is worth recognizing that the permanent GSTV scheme leaves important gaps for certain groups that fall beyond the remit of the program. The income threshold for qualifying for GSTVs will exclude those above the 40th percentile of earners.[20] This means that those earning more than around $2,800 a month—a decent, but hardly comfortable, salary—will face the full force of the GST increase, once the 5-year transition period concludes. While cutoffs have to occur somewhere, it is worth remembering that many households in this income bracket are also in the sandwich generation, who face expenses from supporting their aged (and retired) parents, as well as their young (schoolgoing) children.
One alternative—but admittedly more complicated—scheme for enhancing the progressivity of a GST is to consider a GST where progressivity is built in.[21] This means that the purchase of goods and services that the wealthy are more likely to purchase—fancy cars, high-end property, first-class tickets, and luxury experiences—be subject to a higher rate of taxation.
I am heartened that the government has, in this Budget and over the past year, taken significant steps in this direction. An increase in the additional buyers’ stamp duty by 5 and 10 percent for second and third properties is welcome—if a little blunt since it does not entail any associated minimum value—and PR and foreign purchasers confront even steeper rates.[22] The budget further sharply increases property tax rates for homes with an annual value exceeding $30,000.[23] It remains to be evaluated whether these additional taxes successfully target the wealthy, or whether—much like the ill-fated estate tax—they end up affecting mainly upper-middle and marginally high-income households, while the ultrarich retain their ability to shield themselves from progressive taxation. It will also be valuable to revisit, in a few years’ time, whether such taxes move the needle in terms of freezing or reversing rising income and wealth inequality in Singapore.
There are other tools for raising revenue
The other question that inevitably arises is: are there no alternatives to a GST hike? Mr Speaker, I will submit that there are.
My colleagues at the Workers’ Party have, in the spirit of providing credible alternatives, worked out several additional levers where, if deployed, could stave off the need to increase GST. We have worked out the revenue possibilities for each of these levers, which could amount to close to the $3.66 billion expected from the GST hike.
One lever, which we call the corporate tax lever, accepts the broad premise of the OECD-led effort to implement a global minimum corporate tax rate of 15 percent,[24] known as BEPS.[25] We assume a reasonable loss of corporate and personal income taxes arising from the agreement—of 20 and 10 percent, respectively—and full compliance of the corporate tax rate to the new global minimum for multinational corporations (MNCs), which amounts to a two-thirds increase in the effective rate from current levels.[26] However, we allow small and medium enterprises (SMEs) to retain their current 3 percent effective rate.[27] This could generate $3.45 billion.[28]
Another lever, which we call the wealth tax lever, puts asset taxation at the front and center of revenue generation. We adopt the government’s current threshold of 10 years required for land leases to be classified as “permanent”—above which sales receipts must be fully channeled into reserves—and extract the first 9 years for recurrent expenditure, while directing the remainder into reserves.[29] We also incorporate the government’s proposed property tax changes, expected to generate an additional $380 million.[30] Finally, we follow the net wealth tax tiers that have previously been proposed to this House, to derive revenue from this channel, under very modest recovery assumptions. This could generate $3.70 billion.
A third lever is the reserves contribution lever. The Worker’s Party has spoken about this possibility extensively before. To reiterate: we can reduce the share of reserve interest income that we send straight back to reserves—currently held at 50 percent—to a lower-but-still-respectable 40 percent.[31] Importantly, this does not constitute a draw on the reserve stock—the level of reserves will not go down—but merely represents a reduction in the rate of accumulation of reserves.[32] This alone could generate $4.31 billion.[33]
A final lever, which we call the externalities lever, entails increasing so-called “sin” taxes—those levied on gambling, alcohol, and tobacco—as well as carbon-generating activities, for which we apply an increase to $80 per ton, but channel only half of revenues toward mitigating the transition and encouraging the adoption of green technologies.[34] We limit the increases along each channel to no more than 28 percent, which is the planned rate of increase in the GST.[35] These additional increases would allow us to limit an increase in the effective corporate tax rate to just 8 percent.[36] This could generate $3.65 billion.
Of course, I would be the first to admit that the government, with its army of ministry analysts and superior access to data and information, would be able to fault some (or all) of the assumptions underpinning these scenarios. What, then, is the purpose of this exercise?
First, it is to put on the table the tantalizing possibility that we can, in fact, choose not to raise GST, by adjusting some of the other levers available to us. We have deliberately worked out the math behind these alternatives such that any single one of these levers would be sufficient to meet the revenue needs that GST would offer. Second, we also show that one could even mix and match among the proposals, to craft a revenue mix we can accept. After all, the math suggests that any single one of the levers we propose would be sufficient to fill the GST hole, much less alternative permutations and combinations of them. Third—and perhaps most importantly—it is to underscore our desire to engage in a good-faith debate on whether there are genuine options available to us in meeting the revenue hole, other than raising GST. The most important thing is for us to cease stating, “I wish,” but to start saying, “I will.” We must instead consider nothing impossible, and treat possibilities as probabilities.[37]
On expenditure
Allow me to end with a few quick reflections on the expenditure aspects of this year’s budget.
In last year’s Committee of Supply debate, I flagged the importance of elevating our national R&D expenditure,[38] to shares closer to the most innovative nations in the world. I am happy to hear that the government will continue to increase public R&D spending. I will only add three quick points. First, more can be done, given how we still lag the OECD average for total national R&D spending. Second, this should ideally be directed toward downstream activities—the “D” part of R&D—and hence the stated goal of elevating local firm R&D is the right one. And third, the guiding principles in the BEPS agreement mean that one acceptable way to lower the effective tax rate for our SMEs—and to continue attracting foreign investment—is to do so with R&D tax credits.[39]
I am likewise heartened to hear of the continued efforts to improve the lot of our low-wage workers. The Workers’ Party has repeatedly called for the rollout of a minimum wage for Singaporeans, and the Local Qualifying Salary (LQS) effectively institutes such a floor for a significant majority of Singaporeans.[40] The Progressive Wage Model (PWM) is also being rolled out to additional sectors. I am left to wonder if the Ministry has evaluated how many workers still remain uncovered either by either the LQS or PWM stipulations, and if there is room to increase the LQS to a gross amount of $1,600, which roughly corresponds to a take-home salary of $1,300.
Finally, I believe that the Small Business Recovery Grant, targeted at SMEs and in the sectors hit hardest by COVID-19, is a shining example of how fiscal policy should be timely, targeted, and temporary. Many of my residents who are small-business owners themselves have expressed their appreciation for the support government has offered during this time, although some have also expressed frustration at the paperwork entailed. Simplifying the qualification process would surely be welcome.
Concluding thoughts
Mr Speaker, there is much to like in this Budget, which embeds many elements that I believe will help improve both efficiency as well as equity in our economy. But its stubborn reliance on raising the GST—in spite of alternative, more progressive forms of revenue generation—is the fundamental reason why I cannot lend it my support.
[1] The official initial MTI estimate, issued around the start of 2021, was between 4–6 percent. This was subsequently revised to 6–7 percent in the third quarter.
[2] EIU (2021), Worldwide Cost of Living 2021, London: Economist Intelligence Unit.
[3] The official MTI estimate currently projects real growth of 3–5 percent for 2022.
[4] This final hike followed two prior postponements, in 2015 and 2017, due to concerns that it could trigger an economic contraction. This long-delayed increase, from 8 to 10 percent, finally came to pass in 2019.
[5] Lubik, T.A. & K. Rhodes (2019), “A Closer Look at Japan’s Rising Consumption Tax,” FRB Richmond Economic Brief 19-10: 1–5.
[6] MTI (2009), “Singapore Economy Grew 1.1 Percent in 2008,” Press Release, Ministry of Trade and Industry, Feb 26.
[7] In 2007, global growth was 4.3 percent, while in 2008, it slowed to 1.9 percent.
[8] IMF (2022), World Economic Outlook Update: Rising Caseloads, a Disrupted Recovery, and Higher Inflation, Washington, DC: international Monetary Fund.
[9] Retail sales grew by 10.7 percent in 2021, and fixed asset investments by 11.8 percent. See MTI (2022), “MTI Maintains 2022 GDP Growth Forecast at ‘3.0 to 5.0 Per Cent’,” Press Release, Ministry of Trade and Industry, Feb 17.
[10] This, in turn, arguably means that stimulus effects of government expenditure (especially deficit-financed spending) will be offset by corresponding increases in private sector saving. The effect is known in the literature as Ricardian equivalence. The idea was formalized in Barro, R. (1974), “Are Government Bonds Net Wealth?”, Journal of Political Economy 82(6): 1095–117. The empirical evidence is more mixed and contested, but even if perfect neutrality is unlikely, it is difficult to argue that tax changes have zero influence on consumption and investment. See Bernheim, B.D. (1987), “Ricardian Equivalence: An Evaluation of Theory and Evidence,” NBER Macroeconomics Annual 2:
[11] In the short run, the opposite could also happen. In particular, in a desire to front-run the tax increase, could also spend more before 2023. However, as the experience of Japan shows, the collapse in consumption thereafter may be even more severe.
[12] MAS (2022), MAS Monetary Policy Statement—January 2022, Singapore: Monetary Authority of Singapore.
[13] Among the various methodologies for inflation forecasting—models based on macroeconomic variables, implied breakevens from financial markets, and expectations based on surveys—surveys tend to have the best track record, followed by markets. See Ang, A., G. Bekaert and M. Wei (2007), “Do Macro Variables, Asset Markets, or Surveys Forecast Inflation Better?”, Journal of Monetary Economics 54(4): 1163–212. Currently, both 10-year breakevens from Treasury Inflation-Protected Securities as well as forecasts from the Survey of Professional Forecasters placed U.S. inflation at around 0.5 percent higher than the 2 percent nominal target for the Federal Reserve. Singapore does not have an equivalent survey for the long run, but the latest survey of short-run forecasts has CPI inflation at 2.1 percent, up from 1.7 percent in September. See MAS (2021), Survey of Professional Forecasters: December 2021, Singapore: Monetary Authority of Singapore.
[14] Choo, Y.T. (2022), “Give Economy More Time to Manage Costs Before GST Hike, Economists Say,” Straits Times, Jan 16.
[15] This is based on public remarks at a seminar on Budget 2022 by Ang Yuit, vice-president of the Association of Small and Medium Enterprises.
[16] Fullerton, D., J.B. Shoven & J. Whalley (1983), “Replacing the U.S. Income Tax with a Progressive Consumption Tax: A Sequenced General Equilibrium Approach,” Journal of Public Economics 20(1): 3–23.
[17] Cnossen, S. (1982), “What Rate Structure for a Value-Added Tax,” National Tax Journal 35(2): 205–14.
[18] Krussell, P., V. Quadrini & V. Rios-Rull (1996), “Are Consumption Taxes Really Better Than Income Taxes?”, Journal of Monetary Economics 37(3): 475–503; Nishiyama, S. & K. Smetters (2005), “Consumption Taxes and Economic Efficiency with Idiosyncratic Wage Shocks,” Journal of Political Economy 113(5): 1088–115.
[19] Chen, J.T. & N. Krieger (2021), “Revealing the Unequal Burden of COVID-19 by Income, Race/Ethnicity, and Household Crowding: US County Versus Zip Code Analyses,” Journal of Public Health Management & Practice 27: S43–S56; Daly, P., C. Brassard, J. McCaughey, R. Ng, L. Kathiravelu & B. Horton (2021), “The Social and Economic Impacts of COVID-19 Mitigation Measures on Citizens and Permanent Residents During the Circuit Breaker Period in Singapore,” NTS Insight IN21-02, Singapore: RSIS Centre for Non-Traditional Security Studies; Grewenig, E., P. Lergetporer, K. Werner, L. Woessmann & L. Zierow (2021), “COVID-19 and Educational Inequality: How School Closures Affect Low- and High-Achieving Students”, European Economic Review 140: 103920.
[20] The current threshold is $28,000 in assessable income. In the coming Budget, this will be revised upward to $34,000. The payouts will also increase, from $300 and $150 for those with homes of annual values below and above $13,000, to $500 and $250.
[21] The general concept is that of progressive consumption taxation. See, for example, Carroll, R. & A.D. Viard (2012), Progressive Consumption Taxation: The X Tax Revisited, Washington, DC: American Enterprise Institute.
[22] Ng, H.S. (2021), “Singapore Announces New Property Cooling Measures: Higher ABSD Rates, Tighter Loan Limits,” Today, Dec 16.
[23] Specifically, the rates of between 4–16 percent will increase to between 6–32 percent.
[24] According to the government, the effective corporate tax rate averages between 8–10 percent for MNCs, and around 3 percent for SMEs. See Hansard (2021), “Detailed Breakdown of SMEs and non-SMEs’ Corporate Income Tax Collections,” 95(31): Jul 5. We assume that the effective MNC corporate tax rate is 8 percent, and allow it to double.
[25] The Base Erosion and Profit Shifting (BEPS) project . See OECD & G20 (2021), Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy, Paris: Organisation for Economic Co-operation and Development.
[26] Technically, the GloBE rules in BEPS would only apply to MNCs with revenues above €750 million in revenue. As the OECD has noted, however, countries are free to apply the rates to all MNCs.
[27] Specifically, we assume the departure of firms and redirection of taxable income to their point of consumption is such that the corporate tax revenues decline by 20 percent (of the 2020 take), while personal incomes taxes (arising from job displacement as a result) also declines by 10 percent. These losses could derive from firm relocation (and associated job destruction). But losses are likely to be limited by the global nature of the agreement (141 countries and jurisdictions have signed on), which would in turn limit the relocation choices available to corporates; set against this is the headquarter nature of operations of MNCs here, which may give rise to greater-than-average base shifting.
[28] Others have pointed out how BEPS could well result in net gains in tax revenue. See See, S. (2022), “BEPS 2.0 Unlikely to Hurt Competitivenss, May Net Singapore Gains in Tax Revenue,” Business Times, Feb 28.
[29] The government applies 10 years as a cutoff for when land sales receipts are either channeled fully into operating income, or into reserves. See Hansard (2017), “Classification of Government Revenue Earned from Government Land Leases,” 94(44): Apr 3.
[30] Alternatively, we would prefer introducing an additional tier for property valued above $5 million, which would amount to around $20,000 AV, as had been previously proposed by the Workers’ Party; see Hansard (2021) 95(40): Oct 5. However, there are insufficient data to credibly compute the rates necessary to generate equivalent revenues under such a scheme.
[31] Furthermore, if programs for which expenditures are spent on are well designed, they can help us reduce the need for discretionary government spending. For example, redundancy insurance will obviate the need to provide as much idiosyncratic income support during crises.
[32] Although we do not discuss this point, global interest rates do remain sufficiently low as to potentially justify debt issuance, instead of further reserve liquidation, to finance certain expenditure needs. This is currently the case for SINGA bonds, which are dedicated solely to physical infrastructure. As previously argued, SINGA could benefit from a broader definition of infrastructure that would encompass investment in human capital. This, in turn, would reduce some of the spending needs currently in the budget, especially with regard to education expenditures. See Hansard (2021) 95(29): May 10.
[33] The breakeven for around $3.67 billion is 8.5 percent of the net investment returns contribution (NIRC), which allows for the possibility of underdrawing by up to 1.5 percent of the NIRC in any given year.
[34] This is the upper bound of the government’s proposed amount by 2030, and around $16 shy of the midpoint of the range, of $58–133, that the Workers’ Party had previously proposed during the Climate Emergency motion, debated in 2020 in the House. The revenue from carbon taxes will, presumably, eventually whittle down to a negligible amount, during which marginal increases in the effective corporate tax rate could continue to fill the budget hole left behind.
[35] The two percentage-point hike in GST amounts to a 2/7 increase, or around 28.5 percent.
[36] Unlike the corporate tax lever, we assume here that there is no change in the
[37] Dickens, C. (1849), David Copperfield, London: Bradbury & Evans.
[38] Hansard (2021), “Spending Wisely on National R&D,” 95(23): Mar 2.
[39] Djankov, S. (2021), “How Do Companies Avoid Paying International Taxes?”, Realtime Economic Issues Watch, Sep 3.
[40] While the coverage of the LQS will not extend to firms that do not hire foreign workers, in the absence of barriers to the movement of low-wage Singaporean workers between employers (there are no segmented labor markets) and low involuntary unemployment, the labor market will likely bid up wages for all workers at this low end as well, making the floor close to universal. However, implementation details differ in practice. The LQS will require monitoring and regulation of firms by MOM to ensure compliance with its stipulations. As a consequence, it may be subject to abuse by employers that take advantage of informational asymmetries (workers being unaware of their outside options) and market power (firms being one of few employers in a given area that hires workers of a particular nature) to hire locals below the LQS floor (this may be minimized by relying on real-time CPF submissions). It could also invite gaming of the system, as firms hire and fire foreign workers to free themselves from the LQS, so long as there is a lag in LQS enforcement. In contrast, a well-publicized minimum wage can be more self-enforcing, as it restores bargaining power to workers at the lower end of the income distribution.