The Untimely Demise of the Goods and Services Tax (GST) in Malaysia: A Postmortem and the Way Forward^*

The Goods and Services Tax (GST)¹ is routinely prescribed by the World Bank and the IMF as a panacea for all economies facing revenue shortfalls, regardless of the underlying reasons for such shortfalls. As of June 2023, 175 countries and territories in the world had a GST, or some variation of it (Caragher 2023). The IMF has sponsored two influential books on the many virtues of this tax (Tait 1988; Ebrill et al. 2001) and the issues associated with it. However, what is not immediately apparent is that in many countries what began as a VAT (or GST) has evolved into a structure far removed from the pure version espoused, although they are assumed to be like the original. In the process, many of the virtues of the tax were lost, while existing issues were magnified, and new complications were created. And while the GST works best with a single rate on a broad tax base, in many countries it is implemented with more than one rate on a base narrowed by numerous exemptions and zero-rating (Ebrill et al. 2001). Exemptions complicate the processing of refund claims and auditing, and disrupt the credit trail that enables the self-policing function. Zero-rating, on the other hand, brings in no revenue and leaves the compliance costs on businesses unchanged. The challenges faced by countries in implementing the GST (or some far-removed version of it) are to be found only as case studies buried in academic or professional journals.

In Malaysia, the GST was sold to a government that had accumulated continuous deficits and government debt through years of irresponsible fiscal management, corruption, and wastage of public funds (Narayanan 2012). Despite generating substantial revenues in a short time, the GST proved to be unpopular among small and medium enterprises (SMEs) and the public at large. Dissatisfaction with the GST was so widespread that an election campaign promise made by the opposition coalition to abolish the tax loosened one of the major planks in the base of the ruling government and ensured its defeat in the general elections of May 2018.

Public backlash against a GST is not uncommon and several governments that introduced it have lost power in the aftermath. Governments in at least four countries, Malta, Grenada, Ghana, and Belize (British Honduras) abolished the GST, although all reintroduced it at a later stage, after correcting for the earlier weaknesses (Keen 2009). Interestingly, Hong Kong decided against implementing a GST, after months of intense debate from July 2006 to December 2006.

The pressure to revive the GST has increased in Malaysia, as the nation's budget deficit continues to rise and government debt reaches unprecedented levels. It is, therefore, timely to discuss the reasons for the introduction of the GST, its failure, and the factors to consider if the GST is to be reintroduced. Such a discussion will make a two-fold contribution. First, understanding why the GST proved to be so unpopular will help avoid the same pitfalls if, and when, it is reintroduced. Second, the Malaysian experience may have a wider application to countries that are facing similar difficulties in implementing their GST, or those contemplating introducing it.

Popular public finance texts, largely from the United States, do not provide a detailed explanation of how the GST works. This is not surprising since the United States is a major Western country that has not adopted the GST. We thus provide a brief explanation of the operation of the GST so that readers unfamiliar with the tax will better understand the subsequent discussion on its strengths and weaknesses.²

The production of a good or service goes through several stages before the final product or service reaches the consumer. If, say, the production process is divided into three stages---manufacturing, wholesaling, and retailing---value is added at each stage, through packaging, improvements or other means, before it reaches the consumer. The GST taxes the value that is added at each stage of the production/distribution process.

A simple illustration makes this clear. Let the subscripts $m$⁠, $w$⁠, and $r$ denote the manufacturer, wholesaler, and retailer, respectively, while $P$ is the ringgit value of output, and $t$ is the GST tax rate. To keep the example uncomplicated, it is assumed that the manufacturer uses no inputs to produce the output. (Imagine that the manufacturer's output is the clear mineral water that springs out from her land.) The GST or $Ti$ collected at the manufacturing stage, the wholesaler stage, and the retailer stage, under the credit approach, is given by $Tm=tmPm-0$⁠, $Tw=twPw-tmPm$⁠, and $Tr=trPr-twPw$⁠, respectively. The input of the retailer is the output of the wholesaler and the input of the wholesaler is the output of the manufacturer. More generally, the input purchase made at each stage (except for final consumption) is the sale of the stage before it. This process can be repeated for any number of stages in the production/distribution channel.

The collection of the GST can be illustrated using a hypothetical standard case. The standard case refers to one with a single tax rate applied to all consumption goods without exception. We will then show the impact on revenue when deviations from the standard case occur either through the granting of exemptions, zero-rating, or the imposition of differential tax rates on different stages (or goods).

To simplify the example, we assume that a single commodity from a single manufacturing firm goes through a single wholesale and retail firm before reaching the consumer. However, manufacturers, wholesalers, and retailers can deal in more than one commodity. The tax revenue collected can be generalized using the notational form of the process as shown in Table 1.

The first case represents the standard case where a uniform rate (⁠$ti$⁠) is applied with no preferences such as zero rates or exemptions: thus, the tax rate, $ti=tm=tw=tr$⁠. Also, recall that the manufacturer purchases no inputs. To convert the notation to actual values, assume that the output of the manufacturer, wholesaler, and retailer are valued at RM 200, RM 600, and RM 900, respectively. Using the expression for the manufacturing stage, the tax collected at the manufacturer's stage (or $Tm$⁠) = 0.10 (RM 200) – 0 = RM 20. Similarly, the taxes collected at the wholesale and retail stages are $Tw$ = 0.10 (RM 600) − 0.10 (RM 200) = RM 40, and $Tr$ = 0.10 (RM 900) − 0.10 (RM 600) = RM 30, respectively. The total taxes collected are RM 90 but were obtained from three collection points (RM 20 + RM 40 + RM 30). Equivalently, the total tax collected is given by the expression in the last column, that is, $tPr$ = 0.10 (900).³ This also shows that the consumption-based GST has the same aggregate base as a single-stage sales tax on final consumption. A single-stage tax on the retailer is a retail sales tax; it yields identical revenue to a multi-stage tax like the GST in a closed economy.

Returning to Case 1, note that sellers at each stage collect the tax paid by purchasers and submit it to the tax authority. Buyers at each stage (unless they are customers) receive credit for the tax paid on their purchases, which can then be used to offset the tax due on their sales. This procedure is repeated until the final consumption stage. However, final consumers buy from (say) retailers, and pay the tax due but receive no credit for the tax on their purchases. The tax is, therefore, passed forward fully to consumers. And while the total tax is accumulated from the various points of collection, the onus is on the manufacturer, wholesaler, and retailer to submit the tax collected at their respective points to the tax authority and claim refunds for taxes already paid at the stage before them.

Note that even if a good or service reaches the retailer by skipping any intermediate stage, the final tax revenue remains unchanged. As Case 2 demonstrates, the good reaches the retailer directly from the manufacturer, avoiding the wholesaler but the GST revenue generated is unchanged.

The indirect subtraction method is also referred to as the credit or invoice method. It is known as the former because credit is allowed for taxes paid on input purchases, and as the latter since tax credit on inputs is given only on the production of invoices confirming that the tax has been paid. The need to show invoices to qualify for tax refunds results in the continuous issuance of invoices by each firm in the production and distribution channel to firms making purchases from it indicating that the tax due on purchases has been paid. This creates an invoice trail, a highly touted feature of the GST as it helps identify GST liabilities and verify the tax returns of GST payers. This, in turn, reduces the chances of tax fraud. Nonetheless, since consumers are not required to submit the invoices from retailers to the tax authority, the credit trail ends at the wholesale stage and there is no effective method to verify the sales or tax collectable at the retail stage.

Only taxes on inputs purchased from outside the firms are deductible. This would include taxes on raw materials, components, and parts purchased from suppliers. And while labor services supplied by contractors or purchased from outside the firm are subject to GST and the tax paid on these services is deductible, labor services of employees of the firm do not qualify since their services do not attract the GST. Furthermore, all taxes incurred in the purchase of both intermediate and capital goods may be set against the tax incurred on sales. The deduction (or tax credit) is allowed for the full value of investment goods at the time of purchase. Consequently, the taxable base is reduced to consumption goods only. From a business perspective, the claimed advantage of this type of GST is that it encourages the adoption of capital-intensive modes of production, favors capital formation, and places minimum restraints on working capital requirements.

A weakness of a consumption-based tax is that it is regressive because lower-income households spend a larger share of income on necessities (such as food and transportation) than higher-income households. Thus, a standard tax rate extracts a larger proportion of income (by way of taxes) from lower-income groups as compared to higher-income households. This often results in calls for exempting or zero-rating goods or services used largely by lower-income households from the GST.

Exemption under the GST at best lowers the tax on the exempted good or service---but can also increase the tax on it. This is because the exempted firm (producing the exempted commodity) is assumed to be out of the GST framework and is not required to pay GST. By the same token, it receives no credit for the tax paid on input purchases for that exempted output. The tax embedded in its inputs is therefore passed forward to purchasers in the form of higher prices. If the exemption occurs at an intermediate stage (e.g., wholesale), the wholesaler pays zero taxes as no GST-related transactions are attributed to the wholesaler. The final tax collected, however, will be higher than the tax collected if the wholesaler had not been given an exemption because the tax embedded in the inputs purchased by the wholesaler is passed forward to the consumer (see Case 4 in Table 1).

Granting exemption at the final (retail) stage does not solve the problem entirely either. Since the exempted retailer attracts no GST, refunds on prior taxes paid on inputs cannot be claimed either. The product, therefore, carries a hidden tax element that the consumers bear---albeit less than under the standard rate (see Case 4). Note that exemptions granted at either stage disrupt the credit trail and undermine the self-check mechanism of the GST.

The option to completely relieve a good or service from tax is to zero-rate it, but the zero-rating must occur at the final and not any intermediate stage. This is evident from Cases 5 and 6 in Table 1. A firm selling a zero-rated commodity or service remains within the GST and, therefore, gets tax relief on its sales (because a zero rate of tax is applied) and receives credit for all taxes embedded in its input purchases. However, given its status within the GST framework, it must prepare its returns like all other firms.

Case 5 shows that zero-rating the product at the intermediate (wholesale) stage completely removes the tax on wholesale sales and gives full credit to taxes incurred on inputs. And while zero-rating an intermediate stage relieves the intermediate enterprise from GST (hopefully, for good reasons) it will not affect the final tax. The tax collection process simply bypasses the intermediate stage without affecting the amount of tax collected. Notably, however, unlike exemptions, zero-rating of an intermediate stage does not disrupt the credit trail, does not affect final tax collections, causes no divergence between the statutory and effective tax rates, and does not generate any effect. In contrast, only zero-rating in the final stage frees the commodity from tax as shown by Case 6. Unlike exemptions, zero-rating does not disrupt the credit trail, although extensive zero-rating diminishes the revenue generated from the GST. Even so, zero-rated firms are obliged to apply for tax refunds and therefore their compliance burden is no lighter than non-zero-rated firms.

Multiple rates are sometimes utilized as a device to extend preferential treatment to some goods (say, “necessities”) and discriminate against others (say, “luxury” items). Under the GST the only way to lower or raise the rate on a given commodity is to implement the desired rate at the final stage. Raising or lowering rates in the intermediate stages will simply be “washed away” without giving the desired outcomes as shown by Cases 7 and 8 in Table 1.

To ensure that exports are not relieved from tax and taxes embedded in the intermediate stages, all exports are zero-rated under the GST. This ensures export competitiveness. The way to ensure this is by zero-rating exports. Table 2 demonstrates several examples of zero-rating exports. We assume that there is an exporting and importing country and that there are two stages within each of them. The values of the output of the manufacturer, importer, and retailer are identified by $Pm,Pi$⁠, and $Pr$⁠, respectively. As before, it is assumed that the manufacturer of the exported good purchases no inputs. We further assume that the manufacturer does not export the good directly---preferring to use the services of an exporter.

Case 9 refers to the export that is taxed at the point of entry of the importing country and used directly by agents importing it without going through a retailer. First, due to zero-rating the exporter incurs no GST on the export, yielding a zero-tax liability in column (3). Second, the exporter gets credit for all taxes on inputs incurred in the prior stage (equal to $tPm$⁠). The exported good leaves the country completely free of tax. Zero-rating ensures that the export is not disadvantaged under the destination principle that requires a good to be taxed only at the final destination (or where it is finally consumed), not in the country where it was produced.⁴ This is neither a subsidy nor a favored treatment of exports.

The remaining cases are simply variations of this. Case 10 is an example of an export being taxed at the point of entry of the importing nation and reaching the final consumer there through an intermediary (a retailer in the example), and Case 11 illustrates a situation where the exported good is not taxed at the point of entry in the importing but is taxed at the retail stage.

On the import side, the destination principle requires that imports be taxed in the importing country. To ensure that neither imports nor domestically produced goods are accorded favored treatment, an equal rate of GST is obligatory on both. Columns (4) and (5) show the treatment of an imported good in the importing country. Again, a two-stage process is assumed, although the import could be directly consumed at the importer's stage or be channeled through a further stage (retail, in our example) before reaching the consumer. The treatment of an import would depend on whether it is consumed at the import stage or passes through further stages before final consumption.

If the import is consumed directly or immediately, without going through further stages in the importing nation (Case 9), it must be taxed at the point of importation. Failure to do so leads to a permanent loss of revenue. However, if the import goes through at least one more taxable transaction, it becomes immaterial whether it is taxed at the point of entry or not. This is obvious from comparing columns (4) and (5) of Cases 10 and 11.

Case 10 assumes imports are taxed at the point of entry and that they go through another taxed stage before being consumed. Case 11, on the other hand, assumes that imports are not taxed at the point of entry but also pass through a taxed stage subsequently. The net GST collected in both cases is the same (⁠$tPr$⁠) because if an imported commodity is not taxed at the point of entry, the first taxed transaction will capture all the value-added at previous stages since there are no credits to be claimed.

Interestingly, under-declaring the value of imports will have no net impact on tax revenue because the credit for taxes paid would be correspondingly lower at a subsequent stage.

The greatest danger of distortions in international trade arises in cases when countries use exemptions---even if only in a limited number of cases. To the extent that exempted goods and services carry concealed taxes, these taxes are embedded in exports. Because these hidden tax elements cannot be removed fully, the export commodity enters the world market compromised by concealed taxes.

The ideal GST is one imposed on a broad base (maintained by minimizing goods or services given exemption or being zero-rated) and having a single standard rate. And while it is difficult to find such a structure in practice, the GST in New Zealand gets closest to this theoretical ideal. A measure dubbed the GST's C-efficiency is the ratio of actual revenue over potential revenue obtained from the theoretical GST (a perfectly enforced tax with a uniform rate on all goods and services); the value should be one if compliance is perfect (Cnossen 2022). The C-efficiency ratio is close to one in New Zealand, far above the 2014 OECD average of 0.56 and above all other member countries, except the special case of Luxembourg (OECD Library 2016).

The Malaysian version of the GST was introduced on 1 April 2015, after several false starts.⁵ A single standard rate of 6 percent was imposed on about 60 percent of all consumer goods and services. The narrow base was the result of the zero-rating of 553 categories of goods and services and exemptions granted to another 25 categories (Abas 2018). While zero-rating of goods was probably given at the final stage, it is unclear if the same rule applied to exemptions. Exempting a good that serves as a final consumption item and as an intermediate input complicates clear-cut demarcations. The Malaysian version departed from the theoretically perfect GST; even if implemented efficiently it would, at best, collect only 60 percent of the potential revenue (yielding a C-efficiency ratio of 0.60). In reality, the ratio was probably much lower.

The adoption of the GST was undoubtedly motivated by its reputation as a “money machine” (Keen and Lockwood 2010). This, in turn, stems from several features of the tax. Being based on consumption, the base is broader than that of an income tax because even those with no current income flows must consume some form of good or service. This is particularly relevant to countries with, or are projected to have, an aging population since the revenue from income-based taxes will decline as more people drop out of the workforce. In 2020, more than 7 percent of Malaysians were aged 65 or above, meeting the conventional international definition of an “aging society” (DOSM Portal). By 2044, 14 percent are expected to be over 65 years of age---making Malaysia an “aged society” (The World Bank Group 2020). Since the aged must also consume, the GST will continue to generate revenue in such a scenario. Furthermore, consumption, unlike income, is difficult to hide and individuals who evade or avoid taxes on income must pay taxes on their consumption. Additionally, as the tax is collected from multiple points, it minimizes revenue loss. Finally, at least in theory, the invoice trail acts as a self-policing mechanism against widespread fraud by businesses. An additional advantage of the GST is that, being a consumption tax, it does not discourage savings or investments.

Not surprisingly, voices urging caution in adopting the GST in developing countries have largely been drowned by the influential views of organizations like the IMF when they are approached by cash-strapped economies. Historically, the recommendations of the IMF have consistently stressed economic efficiency and short-term revenue collection goals over other considerations (Itriago 2011). For example, the GST can be problematic when the economy has a large informal sector because not all small businesses can afford to register and invest in compliance. The system is biased in favor of (larger) registered companies and firms not required to register risk being isolated as the former may avoid doing business with them (Chandran 2017). Furthermore, the GST taxes the informal sector indirectly when inputs widely used in the sector are taxed. An economy characterized by weak tax administrative machinery will find the GST too complex to administer. Finally, a country where the bulk of the population earns insufficient income to be caught by the income tax net is not ready for the GST because the burden from the consumption tax cannot be offset (partly) by income tax concessions---as is often done in developed countries.

Nonetheless, for the Malaysian government of the day, the revenue generation aspect proved appealing. The Malaysian economy had experienced continuous budget deficits for over 46 years, between 1970 and 2021, regardless of economic cycles, with surpluses only recorded for five years (1993–97). As of 31 December 2017, just before the ruling Barisan Nasional coalition lost power, the budget deficit stood at 3.7 percent of the GDP, up from 3 percent in 2016. The official government debt was RM 686.8 billion or 50.8 percent of the GDP, but ballooned to RM 885.9 billion or 65.45 percent of GDP when other realized liabilities (such as payments made on behalf of entities unable to finance their government-guaranteed loans) were added. When the lease payments that the government must make for public-private partnership (PPP) projects such as the construction of schools, roads, hospitals, and police stations are factored in, the federal government debt and liabilities amounted to RM 1,087.3 billion or 80.3 percent of the GDP (The Star2018).

Wasteful expenditures and widespread corruption were estimated to cost the economy about RM 10 billion annually or 1–2 percent of GDP every year (Malaysia 2010) while traditional sources of tax revenue were performing poorly. The narrow income tax base was further eroded by populist measures disguised as pro-poor initiatives that have periodically raised the number of tax reliefs and, occasionally, the taxable income threshold. For example, the budget of 2010 raised the personal relief for insurance and Employee Provident Fund contributions from RM 6,000 to RM 7,000 for government servants, while the personal tax relief was raised from RM 8,000 to RM 9,000. The last measure alone removed an estimated 100,000 taxpayers from the taxable income group (Narayanan 2012). Claimed as moves to lighten the burden on low-income groups, it rang hollow as the poor do not pay income tax. It was recently estimated that only 16.5 percent of Malaysia's 15 million workforce (or about 2.475 million individual taxpayers) were subject to individual income tax (Ministry of Finance 2020). The Inland Revenue Department has only about 9,000 people with more than RM 1 million in annual income in its register (Yeap 2019)---a curiously small number given the ubiquity of luxury cars and homes.

Similarly, the corporate tax rate evolved from one flat rate to several rates based on the type of company. The standard rate has been reduced over the years to keep Malaysia attractive to foreign investors. The rate dropped from 40 percent in the late 1980s to the current (2021) rate of 24 percent.⁶ The corporate tax base is also narrow; at the end of 2017, 62.4 percent out of 1,251,190 companies were registered with the Inland Revenue Department, and only 7.8 percent or about 60,898 companies were subjected to tax (Ministry of Finance 2020). Thus, only about 4.9 percent of all companies in Malaysia were being taxed.

Trade liberalization also adversely affected revenue from import levies. In 2013, import taxes accounted for a mere 5 percent of tax revenue and the share was estimated to fall further to 1.5 percent in 2014. The existing sales and services taxes, on the other hand, covered only a limited range of goods and, together, generated 10 percent of total tax revenue in 2013 and the share was estimated to remain unchanged in 2014 (Ministry of Finance 2013). In comparison, the GST accounted for 22 percent of the tax revenue of the Philippines in 1988, 21 percent of Thailand's tax revenue in 1992, and 19 percent of Indonesia's tax revenue in 1985. Singapore's GST with a low standard rate of 3 percent had already generated 8.4 percent of its tax revenue by 1994 (Ebrill et al. 2001).

The impact on the economy from the 1Malaysia Development Berhad (1MDB) scandal was just coming to light. Set up as a Malaysian strategic development company, wholly owned by the Minister of Finance (Incorporated), it was suspected of money laundering, fraud, and theft. The U.S. Department of Justice filed a lawsuit alleging that at least US$ 3.5 billion had been stolen from the state-owned fund. These irregularities received much attention in the media (Hope and Wright 2015). The exact financial implications of the debacle remained unclear at that stage.⁷ With mounting government debt and sluggish growth of revenue from existing taxes, the revenue-generating potential of the GST held great appeal as a quick-fix solution without the need to urgently address wasteful public expenditures or plug leakages.

The initial revenue outcomes of the GST were impressive; it generated a revenue of RM 27 billion over the nine months of its implementation in April 2015 and accounted for 16 percent of tax revenue. In the full year that followed, GST revenue rose to RM 41.2 billion (or 24.3 percent of tax revenue), and it was estimated to yield RM 41.5 billion by 2017 and RM 43.8 billion by 2018 and account for 23 percent of tax receipts in both years (Ministry of Finance 2018). The government was pleased with its adoption of the GST as the coffers were filling up. But businesses, especially smaller enterprises, and the public were not.

Several factors worked in tandem against the GST, but what appealed the most to the widest audience was the claim that it was fuelling the rise in the cost of living. While there is no empirical evidence anywhere to suggest that a GST per se triggers inflation, the opposition's claims appeared to reflect the lived realities of the electorate. We discuss this aspect subsequently, but other potent but less publicized factors were already undermining the GST.

The GST is a tax that was developed in European countries and presumed the ability to administer a sophisticated tax. But the heightened need for revenue and the IMF pushing the tax as a panacea for generating and stabilizing revenue resulted in the implementation of the tax without the requisite supporting administrative infrastructure. In particular, the smooth functioning of the tax refund system is critical to the operational efficiency of the GST. In practice, firms send the taxes owed on their output to the tax authority and subsequently submit claims for the refund of taxes already paid on their inputs with supporting documentary evidence. The tax authority authenticates these claims before refunding the difference. The authentication must occur within a reasonable time frame and firms must receive their refunds promptly to avoid cash-flow problems. The GST Act provided that all GST collections should be placed directly into the Fund for GST Refunds to enable the payment of refunds within 14 working days from the receipt of the GST Return (GST-03) filed electronically (Act 762 2014).

However, the “refund mechanism” functioned poorly, leading to unhappiness and hardship among businesses within the GST framework. It was revealed that GST refunds totalling RM 19.4 billion had remained unpaid to 121,429 companies as of 31 May 2018 and this included the biggest conglomerates in Malaysia. The reason was that the GST collections did not go to the Fund for GST Refunds, as provided for under Section 54(2) of the GST Act 2014, but was diverted to the Consolidated Revenue Account and used for other purposes (Lim 2019). Ironically, alongside this, the Auditor General discovered that the tax authorities had overpaid refund claims, to the tune of RM 4.38 million, between January and May 2019 (Zainuddin and Kaur 2020).

Scant attention was given to tax evasion possibly because the GST was claimed to minimize the problem. But evasion occurs because tax collected by final sellers cannot be verified since receipts from all retail consumers are impossible to collect. Final sellers and buyers can therefore collude for mutual gain. The final seller makes an unrecorded sale to the consumer by waiving the GST on the final sale. This spares the consumer from the tax and allows the seller to understate the final sale. Doing so lowers the amount of tax the retailer must forward to the authority. The revenue loss is borne by the government. Evasion can also occur by wrongly classifying untaxed and taxed goods or making fraudulent refund claims.

It was estimated that GST revenue lost to evasion each year in the United Kingdom had risen from 8 percent to 15 percent of GST revenue in the last decade. Similarly, large losses due to evasion were also reported in Canada (Berhan and Jenkins 2005). Increased auditing raises administrative costs and imposes higher compliance costs on audited firms. The total GST compliance and administration costs incurred by Northern Cyprus (in 2003) and Bolivia (in 2002) were estimated at 1.50 and 1.55 times the total budgetary expenditures to administer their entire domestic tax systems, respectively. In both countries, the compliance and administration costs of the GST refund schemes exceeded 5 percent of the total revenues collected by the GST (Berhan and Jenkins 2005). In Singapore, it was reported that at the end of 2019, more than 300 GST-registered businesses were suspected of being involved in missing trader fraud with an estimated revenue loss of S$ 450 million (RM 1.39 billion) (Singh 2021).

The Malaysian GST was not around long enough to encourage widespread evasion, but had it survived, we would certainly not be in a position to handle it. Over the short period from April 2015 to the end of 2017, more than 5,000 companies were reportedly engaged in GST fraud and revenue exceeding RM 100 million was lost (Bernama2018; Malay Mail2017).

The government overlooked or underestimated the cost of moving to a GST regime faced by businesses, especially the smaller ones. It is commonly held that businesses simply pass the tax burden forward completely to consumers. However, businesses are affected. On one side, passing the tax forward means raising prices and so long as goods have negatively sloped demand curves, this will reduce sales and revenues of businesses in most cases. On the other, there are compliance costs borne by businesses that include the initial expenditure necessary to prepare the businesses to be GST-compliant and the continuous cost of acting as a tax collector for the government without compensation. A recent review of the major studies of GST compliance costs since 1980 found that compliance costs were high and significant in both absolute money terms, relative to tax revenue and the cost of administering the tax in both developed and developing countries. Furthermore, the compliance costs borne by businesses were highly regressive as they fell disproportionately on small businesses (Vishnuhadevi 2021).

In Malaysia, not surprisingly, many small establishments and family-run businesses ceased operations when the GST was first introduced and one or two suicides by small shop owners were also attributed to it (Malaysiakini2015a). In the case of SMEs, a 2016 survey of the Associated Chinese Chambers of Commerce and Industry of Malaysia found that more than 80 percent of its members (comprising SMEs) faced increased financial costs one year after the GST's implementation. Cash flow problems due to GST were reported by 58 percent of the 806 respondents. Members were also unaware of updates on GST-related information issued by the tax authorities. Pointedly, of the 210 respondents facing delays in claiming their input tax, 36 percent said they were not given any reason, nor did they receive any reply to their queries on the delay (Malaysiakini2016). Field studies have confirmed that similar problems plagued other SMEs; apart from the stress and anxiety of the burdensome compliance costs, both tangible (initial, recurring, and non-compliance costs) and hidden (psychological and social costs) were reported. All this was exacerbated by the short implementation period, and the increase in cash flow requirements to fund higher product costs, service costs, and upfront payment of output tax to the tax authorities (Yong, Kasipillai, and Sarker 2018).

As with any consumption tax, the GST burdens the consumer, but it is felt more keenly under the GST than the sales and services taxes because of its broader consumption base. The GST not only raises prices (lowering the purchasing power of consumers) but also extracts a larger proportion of current income from lower-income groups as compared to higher-income groups. This regressive effect was conceded by the World Bank after an extensive survey of developed and developing countries (Ebrill et al. 2001). Moreover, while the GST effectively taxes consumption goods consumed by all income groups, it taxes less completely more sophisticated services like finance and insurance largely consumed by higher income groups (Henderson 1985). The regressivity is best addressed by utilizing the increased revenue generated by the GST to fund pro-poor expenditures. Instead, the bulk of the GST revenue was allegedly dissipated by paying up the debts incurred by the 1MDB fiasco (Sipalan 2018) and giving goodies to a bloated civil service (an important vote bank). In November 2015, barely seven months after the introduction of the GST, and when large sections of the population were struggling under the burden of the GST, the Prime Minister announced an annual pay rise, a minimum retirement fund of RM 950, and a minimum salary of RM 1,200 for the 1.6 million civil servants, claiming that he was redeeming his promise of “returning to the people” revenue collected through the GST (Nazari 2015).

Empirical research on the GST found that it only triggers inflation when it is introduced in economies that were already predisposed to inflation due to other factors like constraints on the supply of goods and services or an easy money policy (Tait 1988, 1990). A detailed study of countries that successfully minimized the one-time impact on the general price level like New Zealand, (South) Korea, (West) Germany, the United Kingdom, and Ireland showed that they either designed the GST to be revenue-neutral (i.e., raise roughly the same revenue as the tax it replaced) initially and/or introduced auxiliary measures to neutralize the pressure on prices. These measures included temporary price and wage freezes, education campaigns designed to prod consumers to act against unjustified price increases, penalties for unjustified price hikes, and well-designed price monitoring mechanisms. Consumer groups played a key role in moderating unjustified price increases. In countries like Denmark, where scant attention was given to similar measures, prices did rise beyond expectations (Tait 1988).

Another insight from the study was that small changes in the consumer price index (CPI) can mask substantial increases in prices across commodity groups. To illustrate, (West) Germany had implemented a GST designed to be revenue-neutral and had auxiliary measures to limit price increases. The one-time rise of the CPI was just 0.4 percent, but the prices of services rose substantially, with gas prices rising by 9.9 percent, electricity by 6.3 percent, and public transportation by 5.2 percent.

Malaysia failed to adequately consider the impact of the GST on prices. The GST was not revenue-neutral and was never intended to be. Indeed, generating urgently needed funds was the overarching impetus for the new tax. The GST covered about 60 percent of all the goods and services in the basket of goods in the CPI (Tay and Amarthalingam 2018) and while the abolition of the existing sales and service taxes resulted in an estimated loss of revenue of RM 17.2 billion (in 2014), the GST yielded an estimated RM 41.5 billion in 2017 (Ministry of Finance 2018). Consumers, therefore, forked out RM 24.3 billion more for goods and services in 2017 than they did in 2014, and the impact on the average consumer was not hard to see.

Also, the GST was implemented at a time when the Malaysian economy was already facing pressures on the price level. The major upward pressure on the price level was caused by the moves to cut subsidies on basic goods such as cooking oil, flour, sugar, and fuel. The subsidy cuts for vehicle fuel, sugar, and cooking gas took effect at midnight on 15 July 2010. The prices of both grades of vehicle fuel rose by RM 0.05 per liter, liquefied petroleum gas rose by RM 0.10 per kilogram, and sugar prices jumped by RM 0.25 per kilogram. Prices for cooking gas increased by RM 0.10 per kilogram resulting in the price of the standard 10 kg cylinder of cooking gas costing RM 18.50, up from RM 17.50 (Malaysiakini 15 July 2010). More aggressive subsidy-cutting measures followed over subsequent years, despite general rising prices, as pressures to rein in the chronic budget deficit intensified. Subsidy cuts to electricity saw an increase in electricity tariffs and public transportation. In January 2014, just a year preceding the introduction of the GST, higher electricity tariffs came into effect with a 15 percent increase in Peninsular Malaysia and a 17 percent increase in Sabah and Labuan (Shagar 2013). Water rates were increased in Johor in August 2015, soon after the GST came into effect (Malaysiakini2015b), and Penang followed suit by announcing an increase in its water “surcharge” in 2018 (The Star2017). Since the country relies heavily on imported food items like chillies, vegetables, fruits, and meats (Sim 2019), the weakening ringgit raised the cost of all imports, including foodstuffs.⁸ If not for a softening of oil prices in the wake of the GST, the average Malaysian consumer would have been completely overwhelmed. While subsidies needed to be rolled back, it was unwise to do it when the GST was being brought in.

The government failed to appreciate fully the impact rising prices would have on the psyche of consumers and took little or no measures to allay that concern. Neighboring Singapore, for instance, instituted measures to ease the GST by cushioning its shock on real incomes (Singapore National Library Online 1994) but the lesson was lost on Malaysia. Also, in many developed countries, taxing individuals on both income and consumption after the GST was avoided by reducing the income tax rates (James and Alley 2009). Since a large proportion of the working population pays income tax in these economies, the move helped reduce the overall tax burden on a large section of the working population.

In Malaysia, the few measures that were taken were ineffective. For example, it mindlessly aped the move in developed countries by reducing the income tax rates for selected bands of income, effective from the year of assessment 2018; this produced a perverse outcome. The tax rate reduction, curiously, did not apply to the lowest annual taxable income band of RM 5,001 to RM 20,000; instead, the rate for the next three taxable income bands (RM 20,000–RM 70,000) was reduced by 2 percent each, resulting in tax savings not only for the affected income bands but for all income bands above them. The first two (lower) income bands enjoyed an average tax savings of RM 300 and RM 600, respectively, while the third and subsequent six income bands with higher incomes received a tax saving of RM 1,000 each (Ministry of Finance 2017). Even if the tax reduction had been carefully targeted, it would at best have benefitted only the small number of workers who were paying income tax. The large section of workers with (low) non-taxable incomes did not qualify for relief from the income tax side but remained burdened by GST on the consumption end. Instead, lower lower-income households were given small cash hand-outs dubbed Bantuan Rakyat 1 Malaysia but the rising prices quickly eroded their value.⁹ Moreover, leakages in implementation resulted in some payments being diverted to those not needing assistance.

The serious attention required in monitoring rising prices in general and the differential impact of the increase in the general price level on different commodity groups within the CPI were lacking. The fall in the purchasing power of ordinary Malaysians was reflected by a severe decline in retail sales (Lim 2015). Protests against rising prices were dismissed by trumpeting the fact that the CPI increases were small.

The widespread disillusionment concerning the GST was capitalized effectively during the general elections of 2018---the ruling coalition that had held power for over 60 years was kicked out of office and the GST was abolished.

The new coalition government in Malaysia, formed after the general elections of November 2022, is once again facing a cash-strapped position. The reasons are varied; the poor financial situation and debt repayment burden inherited from the Najib-run Barisan Nasional government is compounded by the unexpected expenditures forced by the COVID-19 pandemic, a bloated Federal cabinet, and management boards of government-linked companies designed to sustain the hold of two fragile coalitions that followed after the short-lived Pakatan Harapan government. Added to this is the endemic corruption that burdens the economy; at least RM 10 billion is lost annually due to leakages in the government procurement system alone (The Edge Markets2023). Another RM 5 billion was lost in 2020 due to the smuggling of high-value goods (Ministry of Finance 2021) and a total of RM 57.92 million was paid out as subsidies and incentives to deceased farmers between 2016 and 2018 (Nadeswaran 2023).

More generally, leaks and wastage of public funds have gone unaddressed for decades. The direct federal debt by official reckoning stood at 61 percent of the GDP at the end of June 2022, up from the 50.8 percent figure of 2017. Total debt and liabilities amounted to 82 percent of GDP, up from 65.5 percent in 2017 (Zahiid 2022).

The GST looms, once again, as an attractive solution. Endorsed by economists, and some businesspersons, it was lauded as “a mean and lean machine that mobilizes revenue the best” by the local World Bank representative (Morden 2023). These enthusiasts fail to acknowledge that the GST is not a standard beast and even EU countries, with long experience with the tax, are struggling to find the ideal structure (Cnossen 2022). Once again, this quick-fix solution ignores the wasteful expenditures and leakages that bleed the country.

Significantly, economic circumstances like those that tripped the first run of the GST are currently at play and are likely to sabotage the second attempt if care is not exercised. Rising prices of key food items like chicken, eggs, and cooking oil have spearheaded the general rise in prices and threaten to further impoverish lower-income households, although the huge untargeted subsidy for fuels remains in place.¹⁰

Revenue shortfall and rising prices present opposing forces to reintroducing the GST. Falling revenue makes the tax an attractive option while an economy predisposed to rising pressures is not ideal for a broad consumption tax. Yet the GST cannot be rushed; several factors must be considered before the tax is presented to the public again.

The new Prime Minister (who took office in November 2022), in his role as Finance Minister, recognized the importance of timing while not denying the role of the GST in generating and stabilizing revenue. He declared that the tax would be inappropriate in the current environment where a significant proportion of the population was struggling with low incomes and the CPI rise of 3.2 percent over the last year (2022) was paralleled by a 5.8 percent increase in food prices. He promised, instead, to address leakages in subsidies and other forms of wasteful expenditures (Soo 2023). Aggressive initiatives to curb corruption, wastage, and leakages in the public sector will inspire confidence in the government's fiscal management abilities and make future tax proposals more widely acceptable.

Given that the GST is a known beast, the temptation to push it through will be strong but should be avoided. Preparing the ground for the tax is of crucial importance and the New Zealand example offers useful insights.¹¹ The tax should be presented as an effort to make it better than before rather than to accommodate opposing demands. The initiative should begin well before the tax is implemented. In New Zealand, this began with a White Paper on GST that provided a simple technical description of the way the tax will be applied, and the responsibilities required of businesses and the self-employed. This was accompanied by the creation of a lean GST Coordinating Office with a clear objective and a predefined, short lifespan of 18 months. It not only developed and implemented a public education and information program but also coordinated the introduction of the tax. The Office was led by a three-person team comprising a senior partner in a renowned accounting firm (with no political profile), a public figure with experience in consultancy work, and a senior analyst with experience or familiarity with a well-designed GST elsewhere. One key aspect is to gain acceptance for the idea of taxing virtually everything for minimizing the regressivity of the tax. Taxing everything, including “necessities,” will generate more revenue that can be redistributed to the deserving. This must be a promise that is not only delivered but seen to be delivered for the new GST to gain acceptance. This important message must be part of an honest education campaign that precedes implementation.

Despite facing a formidable task, the rise of the GST acceptance rate in New Zealand from 35 percent to 65 percent within two weeks of introduction speaks volumes about how the tasks were performed. The trust and respect gained from informed taxpayers is the cornerstone of successful reform.

To live up to its name of being an “efficient,” distortion-free tax, the GST must utilize a single standard rate imposed on all final consumption, with minimum exemptions and zero rating (Enache 2022). Yet almost everywhere it has been implemented on a narrow base with high rates (often more than one) and with numerous exemptions and zero-rating in the name of reducing regressivity. However, the aggregate of the policy gap (exemptions, reduced rates, thresholds) and the compliance gap (revenue shortfalls due to laps in compliance and implementation) increase distortions and compromise the revenue productivity of the GST (Cnossen 2022).

The fact remains that taxing consumption goods with a single rate is politically controversial---even more reason to keep the rate low, but this is only possible by broadening the consumption base. Arguably, high rates imposed on a narrower base are more likely to be resisted than a single low rate on a broad base with minimum exemptions or zero-rating. In Singapore, for instance, the GST came into effect in April 1994 with a low rate of 3 percent, with an assurance that the rate will stay the same for seven years, despite the expected loss of revenues (Singapore National Library Online 1994). In New Zealand, introducing a revenue-neutral GST initially led to persistent and growing deficits before base broadening measures and a buoyant economy kicked in to generate surpluses (James and Alley 2009).

The base is broadened by minimizing exemptions and zero-rating and not by setting an unrealistically low sales threshold for GST registration. An annual sales turnover of RM 1 million (or more) might be more appropriate than the previously set sales of RM 500,000. In New Zealand, for example, even commodities such as food, children's clothing, books, and medicine, and sectors like government, residential dwellings, local government, and tourism, are taxed. Only domestic rental accommodation and financial services are exempted for practical reasons (Inland Revenue 2021).¹²

The regressive impact of a flat rate tax on consumption must be addressed, but not through extensive exemptions and zero-rating because they are largely futile. Instead, taxpayers should be made to understand that exempting “necessities” creates difficulties in identifying boundaries between taxable and exempt items, leading to revenue leakages and economic distortions. Besides, zero-rating or exempting commodities creates untargeted subsidies that benefit both deserving and undeserving consumers.

The impact of the GST on the cost of living must be acknowledged and measures to soften its blow on the income side should take the form of tax credits to income taxpayers in the lower income bands (and not through rate reductions, that yield greater tax savings to higher income taxpayers), and targeted cash subsidies to non-taxpaying households. In Singapore, such measures were put in place a year before the GST took effect. They included increased grants for lower-income households and increased subsidies for some public services such as health and education. An anticipated rise in GST saw the distribution of Economic Restructuring Shares in January 2003, with households being allocated shares according to their incomes. The shares promised an annual dividend of at least 3 percent from 2004–08 (Singapore National Library Online 1994). In addition, auxiliary measures to minimize price increases (as discussed previously) and systematic monitoring of the CPI and its components should be institutionalized.

Lowering the compliance rate should be prioritized. It is tempting to assume that the GST can be resurrected quickly because of existing registration records. But any move to remove or reduce exemptions and zero rates will call for a revision of existing records as will the raising of the sales threshold. New firms that have emerged in the interim will need assistance to meet the fixed compliance costs. Easy access to low-interest or no-interest financing for small enterprises will help overcome the need for a fresh outlay of funds.

A low uniform rate, with few exemptions and zero rates, will also lower the compliance costs for businesses and reduce tax administration costs. Keeping the registration threshold realistic, as argued earlier, also lowers compliance costs and strengthens the self-enforcing feature of the tax. In New Zealand, compliance costs were lowered by also allowing creative options for accounting such as submitting GST returns on a monthly, two-monthly, or six-monthly basis, or annually with provisional tax payments. The use of cash, invoice, or hybrid methods of accounting for GST further reduced compliance costs (Inland Revenue 2021). Most businesses, therefore, complete their own GST returns without recourse to professional services.

The efficient functioning of the refund system cannot be overemphasized. Delays in receiving refunds can strain business operations, especially for small businesses. Malaysia's GST floundered on this aspect, and it is essential to rectify the error. The refund mechanism so critical for the smooth operation of the GST also opens a wide avenue for tax evasion or fraud. The frequent point of evasion is where final sales to consumers occur (say, the retailer). Because sales of the retailer cannot be counterchecked, the retailer can lower the tax owed by understating sales. Theoretically, a retailer can also collect money from the tax authority by understating sales so that the retail value-added is negative (and no tax is due) and submitting claims for a refund of taxes paid on input purchases. Among the options to avoid such fraudulent practices is to legislate that no instant refunds will be granted in cases where tax credit or input purchases exceeds the tax owed on sales; rather, the credit will be carried forward and deducted from the future tax owed (Harberger 1990). Of course, this mode of cheating is self-limiting as it attracts audits if any firm engages in it regularly.

Even when fraud is detected, auditing is neither easy nor costless. To illustrate, in Singapore, auditing a trader suspected of fraud results in the “freezing” of claims of all traders dealing with that firm, even if they are not involved in the fraud. The Singapore authorities have been unable to resolve this issue for several years and smaller enterprises have borne the brunt of extended audits that delay their input tax refunds. To address the problem, the authorities have come up with a less-than-ideal solution---it shifted the burden of auditing GST-registered suppliers to GST-registered businesses, effective from 1 January 2021 (Singh 2021).

Minimizing preferential treatment not only broadens the tax base but also simplifies the auditing process; zero-rating and exemptions granted to the same firm producing different goods complicate auditing. Avoiding zero-rating also reduces the incentive to misclassify goods to make a gain by exaggerating the share of zero-rated goods relative to non-zero-rated ones. Exemptions provide a similar incentive to cheat; as no refunds are given for taxes paid on inputs used for exempted goods, firms have the incentive to overstate the output of exempted goods and understate the inputs used in their production by assigning those inputs to products subjected to GST, so that credit could be claimed (Harberger 2002).

There are probably countless other ways to commit fraud as the ingenuity of tax avoiders cannot be underestimated. It underscores the need for continued vigilance as there is no floor to the revenue risk (emphasis added) arising from the fraud and loopholes enabled by the tax refund system (cited in James and Alley 2009).

The advantage of a properly designed GST as a stable, efficient, transparent, and effective revenue source cannot be ignored. Malaysia may have to return to the GST at some point in the future, but it is important to avoid viewing it as an immediate and permanent solution to rising government deficits and debt.