This chapter draws on an article prepared for the Group of Twenty-Four, “Political Economy of Tax Reforms,” which was presented at the IMF/World Bank 2019 Spring Meetings (Ahmad 2018a), as well as at an ongoing, joint London School of Economics and Coalition for Urban Transitions research program on financing sustainable development in China and Mexico (Ahmad and others 2020).
Tax design can help address two types of informality that affect emerging market and developing economies. The first relates to general incentives to “cheat” on formal taxes or contributions, facilitated by hard-to-tax smaller businesses interacting with large enterprises, with segmentation of the tax bases. Incomplete information flows can seriously affect the general government revenue position and ability to finance basic services and public investments. The second concerns the popular perception of the informal sector as poor migrants and marginalized groups in metropolitan areas, often living in shanty towns with limited access to public services, such as education or health care, and extremely vulnerable to pandemics, such as coronavirus disease 2019 (COVID-19).
The Relationship Between Informality and Tax Policy
The presence of people willing to work without contracts facilitates informality of the first kind, enabling firms to hide payroll taxes that add to the cost of doing business by disguising transaction volumes, wages, and profits. Informal workers often lack rights to formal benefits, access to public services, and ability to attract credit.
Informality as “cheating” typically arises when firms have both an incentive to evade taxes and the ability to do so. Such opportunities typically originate as well-meaning government measures to provide investment incentives or social benefits through exemptions or preferences in the tax system or segmentation of tax administration between large taxpayers and a residual small taxpayer regime that may be delegated to lower levels of government. The resulting provisions create discontinuities that generate rents, as well as breaks in the information chain. Often, these provisions do not achieve either distributional or investment objectives and seriously hamstring revenue generation. Examples are found in Pakistan over the past 30 years (Ahmad and Mohammed 2018) and in Mexico before its 2013 reforms.
Standard prescriptions to enhance distributional outcomes, or to encourage investments and ease administration, often generate counterproductive outcomes. Redistributional policies may increase incentives to cheat and reduce revenues while generating negative consequences for the informal sector and the poorest groups. Adverse effects include complicating the value-added tax (VAT) with multiple rates or exemptions, as well as a VAT applied only to large taxpayers. VATs that exempt natural resource investments and special economic zones (SEZs) also create incentives to cheat, as seen in the Nigerian petroleum sector (Petroleum Revenue Special Task Force [PRSTF] 2012).
Other complex measures, such as payroll taxation on formal sector workers, in replicating Bismarckian social protection instruments, also increase the cost of doing business and create double taxation of labor, thus encouraging informality. Further, means-tested benefits, including the globally popular conditional cash transfers, generate severe disincentive effects and often create poverty traps. In Chiapas, the poorest state in Mexico, such unintended effects prompted abolition of the Oportunidades,1 the conditional cash transfer program that has been copied by many other countries worldwide.
National Tax Reforms to Discourage Informality
National tax measures affect incentives facing firms and workers, hence informality and revenue potential. The design and administration of a tax system influences the cost of doing business and prices, affecting both producers and households. This in turn affects the distribution of income and spending. Insights from optimal tax theory, as well as the simpler theory of reform, provide directions for welfare-enhancing reforms (Ahmad and Stern 1991) that can be used in conjunction with the structural measures to consolidate tax bases and information generation.
Several countries have triggered structural reforms by shifting from taxes that increase business costs to more efficient sources of revenue, such as the VAT, that do not distort production decisions or tax exports. Policies that raise minimum general revenues while reducing the cost of doing business have centralized the main tax handles. Reform has not always been easy, as in seen in China in 1993/94 (Ahmad, Rydge, and Stern 2013) and Mexico in 2013 (Ahman 2015), and focused on consolidating national taxes, particularly the VAT. Neither addressed the subnational need for own-source revenues, permitting control over rates. Shared revenues are like transfers and do not generate appropriate incentives at the subnational level.
Emerging market and developing economies can err by expecting one instrument, such as the VAT or the corporate income tax, to simultaneously generate revenues, encourage investment, and address income distribution.
Transplanting from Organisation for Economic Co-operation and Development countries a formal social security system financed by payroll taxes similarly adds to the costs of doing business and could thus increase both types of informality—the incentives to cheat, as well as the incentives facing workers to remain in the informal sector. As seen in China (Ahmad 2018b), a tax reform “package” is needed, covering a range of taxes and transfers to offset gainers and losers at subnational levels. The package should include tax measures that minimize distortions and reduce the costs of doing business, as well as public investment and social spending designed to motivate sustainable and inclusive growth.
As part of any reform package, an integrated VAT is critical. A “clean” VAT free of loopholes and exemptions generates complete information on the production chain, thus reducing the ability to hide transactions and informal employment. Integration of the VAT base was a critical element in the 2013 Mexican reforms and helped prevent both perverse incentives and evasion, not just for the VAT but for excises, payroll taxes, and income taxes.
Given that wide-ranging tax reforms generate both gainers and losers, political economy considerations require either joint implementation of taxes and transfers (as with the Chinese 1993/94 reforms) or simultaneous offsetting tax reforms (as with the 2013 Mexican reforms). China and Mexico offer the most significant examples of systemic tax reforms in major emerging market economies in the past three decades, with major revenue and structural consequences. Both country’s reforms were based on introducing or fixing the VAT and offer lessons for emerging market economies in an increasingly global world. As argued in this chapter, reforms in China and Mexico helped reduce informality and cheating by firms, consequently raising the revenues needed to anchor spending for basic needs.
Subnational Own-Source Taxation for Sustainable Development Goals and Addressing Informality
At the subnational level, broad-based access to basic services generates incentives for informal workers to move to the formal sector. Such services constitute the core of the UN Sustainable Development Goals (SDGs) and are also critical in addressing the global pandemic and building back better. Own-source local revenues to finance the SDGs are critical in ensuring accountable policymaking. These own-source revenues are also the basis to access other financing mechanisms, including municipal bonds and capital markets, to ensure adequate infrastructure investments needed for sustainable growth.
Well-designed clean, compact, and connected cities can become sustainable employment hubs and engines of economic growth, provided that adequate local supplies of public goods are available to the entire population. Yet, if most productive activity and employment opportunities are concentrated in a few metropolitan areas (as in Dakar, Guangzhou, Mexico City, Mumbai, or Santiago de Chile), megacities will continue to attract migrants in search of better living conditions. The result is growing informal settlements as well as urban sprawl—the opposite of the clean and compact cities needed for sustainable development.
Badly designed cities, financed by land value capture, contain islands of wealth but also generate urban sprawl, congestion, pollution, and inequalities. Poorer households are often evicted from rapidly appreciating areas and are pushed into slums in the outskirts of town. Furthermore, land value capture often involves eviction through land grabs, facilitated by vested interests, to build low-density villas or golf courses that ultimately pay little tax.
Standard prescriptions concerning property taxation can also be counterproductive and contribute to the increase in informality, typically seen in metropolitan areas such as Cairo, Karachi, and Mexico City. The typical property tax, based on ownership and accurate valuation, further discriminates against poor migrant squatters or tenants who cannot prove residence to access public services. Moreover, enforced titling on communal land can break down social cohesion and the risk-sharing mechanisms critical in countries with weak or no social protection mechanisms.
Property Tax as a Cornerstone of Sustainable Development
A typical model of property taxation, based on ownership titles and accurate valuation, as in the United States, does not work well in emerging market economies. Ownership titles are often unclear, and accurate valuation is difficult to obtain in a timely manner. Even in the United Kingdom, this type of model was abandoned by former Prime Minister Margaret Thatcher. The political economy constraint in countries like India, where the tax was introduced in the late 1800s, is even more problematic—the taxation of fixed- or low-income households living in rapidly appreciating properties that have often not been revalued for decades.
Yet property taxation remains critical for equitable and sustainable development, even if aggregate revenue potential is limited to 2 to 3 percent of GDP and is dwarfed relative to broad-based taxes, such as the VAT. Even if 1 or 2 percent of GDP could be generated locally, as opposed to the typical collection of about 0.25 percent of GDP in many developing countries, including in Latin America and India, this would be significant in financing basic services, because such taxation also helps facilitate access to credit for local investments and infrastructure. If linked to basic services, a beneficial property tax would generate incentives for better governance.
The alternative “beneficial” property tax could be based on simple identification of occupancy, size of property, and location, but with rates linked to basic services. The linkage with basic services is critical in overcoming political resistance, developing a theme suggested by Alfred Marshall (1890). The identification of size, location, and occupancy is straightforward and sidesteps the problems with ownership and valuation. The use of satellite imagery and blockchain technology are promising new methods of administration that can further simplify application. Using simulations for Chinese cities, Ahmad and others (2020) illustrate the effects of a simple area-based property tax yielding 2 percent of GDP. Such revenue can finance basic education, provide poorer households with education stipends, and support social housing. Appropriately designed property taxation allows informal households better access to public services and can play a pivotal role in addressing health crises, such as with the pandemic.
Proposals for a beneficial property tax for Mexico suggest a significant short-term revenue potential of about 1.5 percent of GDP rather than the 0.25 percent of GDP currently generated by the traditional property tax. The beneficial property tax could also lead to reductions in inequality and become an anchor for the post-pandemic “building back better” agenda for the informal workers in metropolitan areas and those who returned home in the lagging southern states (Ahmad and Viscarra 2020).
The Relationship Between Formal Social Protection and Informality
Informality as cheating arises when firms and workers have an incentive to hide transactions from tax administrators and the design of the tax and information system makes it easy for them to do so. Measures to address investment or distributional objectives often create discontinuities that encourage informality.
One such example is the argument by Levy (2008) that inappropriately designed formal social protection systems can raise the cost of doing business and increase incentives for firms to hide profits, turnover, and outlays on labor (see also Antón, Hernández, and Levy 2013). Firms then shift workers to temporary and informal contracts. Both forms of evasion lead to inefficient outcomes and reduce growth. The social protection system thus constitutes, as Levy writes, “good intentions, bad outcomes.”
National Tax Reforms in Mexico to Block Rent-seeking
Levy’s recommendation for Mexico was to shift from high payroll tax on firms (which adds to the cost of doing business) to the VAT, which, when well designed, is neutral to production and exports. However, Mexico, before its 2013 reforms, did not have a VAT capable of generating sufficient revenues to replace the payroll tax. Since the early 1980s, when Mexico introduced its VAT, successive governments aimed to provide incentives for investment, production, and redistribution by reducing rates in sensitive or “border” regions outside the tax-free maquiladora zone and providing exemptions and domestic zero rating.
The base for the Mexican VAT (as well as the income and other major taxes) was further split as the small taxpayer system (REPECOS) was allocated to the state level. With the federal government granting gap-filling transfers to meet state deficits, states had no incentive to pursue hard-to-tax groups. A study for the
Mexican Tax Administration estimated over 90 percent evasion of the REPECOS (Fuentes Castro and others 2011). The net result was that the ratio of non-oil tax to GDP was about 10 percent for general government, or similar to Pakistan— well below the average for Latin American and Caribbean countries or about 15 percent of GDP. The C-efficiency of the Mexican VAT was also not dissimilar to that of the Pakistani VAT, and at about 25 percent was among the lowest in the world (Antón, Hernández, and Levy 2013). An inefficient VAT could not therefore replace the distorting payroll tax.
The Mexican income tax (ISR) also suffered from base erosion. Many beneficiaries of exemptions or preferences were large firms with political connections. Once given, provisions and tax breaks become virtually impossible to remove—a pattern that resonates in other emerging market contexts (and in advanced economies as well). The Mexican government initially followed Latin America’s widespread attempts to implement a gross asset tax, called the minimum asset tax, in Mexico. As with the ISR, the incidence was on the largest taxpayers, who could maintain their preferences, and it did not have much of an effect on addressing the poor revenue collections.
Stand alone reforms to specific taxes had not been successful, presaging the reform efforts to follow. To remove preferences and deductions, governments must confront powerful vested interests. Reform thus becomes complicated if some parties can use state interests to block measures in the senate. In 2007, the Mexican government was not able to overcome senate opposition to address loopholes in either the ISR or the VAT. An indirect mechanism was consequently chosen—a minimum tax in the VAT mode credited against the ISR through a single-rate business tax, Impuesto Empresarial a Tasa Única (IETU), which replaced the minimum assets tax. It worked initially but soon began to be affected by the distortions and exemptions that were the problem in the ISR in the first place.
Passage of the IETU reform was facilitated by a Chinese-style stop-loss provision and a rationalization of intergovernmental transfers to ensure that no state lost revenues because of the reforms. The “reform package” consequently had sufficient support to be approved by Mexico’s congress. Although the single business-tax rate had some disincentive effects, the underlying value-added design did not disadvantage investments as much as a turnover tax would have.
A later attempt in 2010 also failed to close VAT loopholes by providing households with additional benefits through the Oportunidades conditional cash transfer. First, many of the losers from a VAT reform are urban households, and conditional cash transfer provides benefits mainly to poor rural households, making it an inappropriate compensation mechanism for many energy price and tax reforms. Second, the main losers in a VAT reform tend to be firms with vested interests and possibly some states, given Mexico’s revenue-sharing system. The interactions between the VAT loopholes and the provisions in the ISR present formidable incentives to evade taxes and engage in informal activities, but such incentives cannot be addressed by conditional cash transfers to poor rural households.
Other problems with conditional cash transfers became apparent in Mexico and are relevant for countries considering both tax reforms and energy price adjustments. The principal problem is that conditional cash transfers create a poverty trap that encourages informality and reduces mobility. Poverty rates and uptake of Oportunidades rose simultaneously for 20 years in Chiapas. Conditional cash transfers were finally abolished in 2019 (by a left-wing government, although the Peña Nieto administration had attempted a significant overhaul in 2014).
As stressed by some academics, a turnover tax is offered as a third-best option for countries that have difficulty with revenue generation (Best and others 2015). This was never an option in Mexico, because the government knew that such a tax would add to the cost of doing business and would adversely affect Mexico’s competitive position with its trading partners.
The Mechanisms of Tax Evasion by Mexican Firms
Firms use two mechanisms for evading or avoiding taxes. In the first mechanism, firms avoid VAT or ISR by remaining below the annual turnover threshold or by splitting into smaller firms. This is the typical case discussed in the literature and formalized in Keen and Mintz (2004). The revenue losses from such reconstitution are likely to be small, prompting a strong recommendation from international agencies that countries raise the registration threshold to reduce burdens on the tax administration (IMF and others 2015). Yet small traders and small and medium enterprises exercise enormous political power in countries such as Pakistan and are an obstacle to meaningful reforms—unless losses might be shown to offset by a rationalization of income taxes, as in Singapore (Ahmad 2018a). Rationalization, however, requires the establishment of an arm’s length tax administration that does not impose additional costs of doing business, especially on small firms.
The second tax evasion mechanism, and by far the more important element in Levy’s (2008) view, is outright evasion by larger firms, which also have much greater political clout than smaller traders in a more advanced economy such as Mexico’s.
The principal channels for rent-seeking are twofold:
1. Incentives to cheat are generated through tax systems that encourage pricing discontinuities and arbitrage. Such systems may charge multiple VAT rates for the same goods, implement exemptions for border regions, and privilege maquiladoras through creation of SEZs. Together with loopholes in direct taxes, high marginal rates that add to the cost of doing business, and domestic zero-rating, such incentives encourage and reward rent-seeking.
2. Exemptions from the VAT lead to absence of information on transactions, creating silos of tax information within tax groups and across tax instruments. Breaks in the value-added chain and presumptive mechanisms to estimate VAT liability, when used extensively as in Pakistan, prevent the self-policing aspects of the VAT from operating. Such breaks are exacerbated by weak or ineffective administration for components of the tax system. In Mexico, the weakness was the REPECOS system that strongly encouraged informality.
Before Mexico’s 2013 VAT reform, both channels for rent-seeking operated in tandem, as summarized in Figure 9.1.
Figure 9.1 identifies several tax evaders under the pre-2014 regime:
REPECOS and adjusters. The REPECOS tax regime was for small businesses with turnover below Mex$2 million. Few REPECOS firms paid any tax in Mexico. Although the VAT has no legal registration threshold, turnover below Mex$2 million became the de facto threshold. Smaller firms were effectively ignored by the Mexican Tax Adminstration, but this group of taxpayers also includes the “adjusters,” who legally reduce turnover to stay under the REPECOS threshold.
Enanos, or “ghosts.” Many so-called enanos firms are too large to be eligible for REPECOS but pretend to be eligible regardless. Furthermore, because the enforcement of REPECOS is weak, businesses know that if they at least pay some REPECOS dues, the state governments and the Mexican Tax Administration will likely be satisfied. These firms are the ghosts identified by Kanbur and Keen (2014).
Larger firms. As argued by Levy (2008), middle- and large-size firms often hide transactions, turnover, employment, and profits by trading with the enanos, truly REPECOS-eligible small firms, and other rent-seekers. Larger firms can thereby reduce payroll and profit taxes by avoiding the VAT chain.
“Honest” firms. International agencies often assume some firms are not able to evade taxation because they are either run by multinationals or too large to do so undetected. Yet, large firms are often the best-connected “vested interests.” Moreover, multinationals are better able to avoid taxation than most large domestic firms, for example, by moving corporate headquarters to low-tax jurisdictions, such as Ireland or Luxembourg, if not The Bahamas or Panama. However, we maintain the “honest” firm categorization in Figure 9.1 for completeness (the dark blue squares to the right—whereas the “true” characterization is in the white and blue squares immediately to the left).
All possible forms of VAT fraud are greatly facilitated by the existence of the maquiladora tax system, particularly by its lack of transparency. This allows firms to hide and disguise taxable activities, including within the rest of the country, and the ever-widening definition allows more and more firms to take advantage.
Figure 9.2 summarizes the revenue losses created by interaction between the maquiladora system (that is, SEZs) and the VAT. The figure depicts a standard carousel fraud. Businesses that import inputs can pass the input credits to another Mexican firm, which can then export and claim the input credits. This fraud is greatly facilitated by the ability of maquiladoras to operate as part of a group of firms both within Mexico and abroad, with few reporting requirements to the Mexican authorities. Figure 9.2 also depicts a more straightforward export fraud in which a pair of related firms, one a maquiladora, collude to claim an input credit for a transaction that never occurred.
Figure 9.2 depicts the most pernicious fraud as well. Under the maquiladora system, bonded imports do not incur a VAT liability, provided that the transformed outputs are reexported. These inputs or their resulting transformed outputs are then passed to another Mexican company, which sells them in the domestic market without ever having paid VAT on the imported inputs. Customs data show that only an estimated 50 to 60 percent of the inputs imported under this system ever leave Mexico.
A comparison of the profits declared to the Mexican Tax Administration versus those implied by the economic census (which may itself be an underestimate) indicates the possible extent of underdeclarations and their components and support the Levy (2008) hypothesis. The analysis also shows that both the incentives and the ability to cheat exist for firms of all sizes. Small and medium enterprises do not appear to have a monopoly on cheating the tax administration. As firms fail to declare turnover and profits, they push workers into informal contracts to further disguise the scale of their operations. Thus, one type of informality, cheating, thrives on and perpetuates the traditional definition of informality— workers without formal benefits.
The Mexican Tax Reform Package of 2013
The political economy of Mexico’s 2013 package of tax reforms was instructive—with all main taxes reformed together rather than treated separately. Because these reforms were designed to close loopholes, the interactions between a reformed VAT and income taxes were particularly important. In addition to the elimination of policy loopholes, the principal administrative reform was thus to replace REPECOS and the intermediate tax regime by a new Régimen de Incorporación Fiscal (RIF) integrated with the regular tax system. This required small taxpayers below the Mex$2 million threshold to use a simple cash-flow electronic accounting package (provided by the Mexican Tax Administration) and issue electronic invoices.
All taxpayers were accordingly brought under the Mexican Tax Administration’s registration, compliance, and audit. Anomalies in the VAT itself were removed, as were border-region rates, whereas the standard rate was retained. The goal of the reform was to bring the whole value chain under the Mexican Tax Administration’s supervision. The income tax was rationalized, but its rate structure was maintained.
With the full VAT in operation and electronic tracking of invoices, the ability of firms to engage in hidden transactions was effectively eliminated. Mexico thus had little need to maintain the IETU minimum tax, and it was abolished, reducing administrative burdens on taxpayers. The only additional tax introduced was a carbon tax or petroleum excise, above an adjustment of petroleum prices toward world prices.
The political economy of the reform was important, and the losers with respect to one major tax were offset by gainers on others. The reform was supported by all parties, because the package would enhance the basis for long-term growth, minimize the effect on poor households by excluding unprocessed food from the VAT, and create a basic minimum pension that would not generate labor market disincentives or a poverty trap, ensuring that the most vulnerable, especially in urban areas, would not be affected by relative price changes. Furthermore, the reform maintained VAT information flows, because unprocessed food does not enter the middle of the production chain. The only compensatory measure was the basic pension for those older than 65 years who did not have alternative occupational pensions. The typical recommendation to facilitate the adjustment in energy prices through Oportunidades was not implemented.2
Outcomes for the 2013 reforms exceeded expectations. Despite the period of low growth affecting much of Latin America, largely attributable in Mexico to depressed petroleum prices, the 2013 tax reform offset most of the fall in petroleum revenues, raising over 4 percent of GDP by 2016 with no increase in the rates of the major tax instruments.
National Tax Reforms in China
China is of interest because its tax and structural reforms were closely linked. After Deng Xiaoping introduced the Responsibility System in the late 1970s, both total revenues collected and the proportion shared upward with the central government fell precipitously, yet incentives for producers improved. Yet, because China did not have a central tax administration, the total tax-to-GDP ratio fell from 25 percent to about 10 percent by 1992/93 (Ahmad, Rydge, and Stern 2013), despite attempts to encourage local officials to share revenue with the central government (Gao 1995). The central share of tax revenues, which had been about 55 percent of total revenues at the start of the Responsibility System reform, fell to well under 30 percent by 1993 (Figure 9.3) as local governments sought to protect local interests in the face of falling revenue collections.
In the absence of modern fiscal institutions and instruments, the structural changes generated a fiscal crisis. The plummeting tax-to-GDP ratio, as well as the decline in the share going to the central government, seriously compromised the central government’s ability to maintain macroeconomic stability, ensure redistribution, and meet the fundamental responsibilities of a nation-state. By 1993/94, China needed a major tax reform to consolidate the structural reforms initiated by the Responsibility System and to strengthen the central government’s ability to maintain macroeconomic stability, redistribution, and public investment (Ahmad, Gao, and Tanzi 1995; Ahmad, Rydge, and Stern 2013).
The 1993/94 Fiscal Reform Package: Consolidating Structural Reforms
The establishment of the (central) State Administration of Taxation, along with central revenue-raising powers, was facilitated by the introduction of an investment-type VAT in 1994. However, a package of tax administration and policy reforms was needed to minimize losses and share benefits across rich and poor provinces alike. The key elements of the reform “package” include the following:
Preventing losses among local governments through a hold harmless clause guaranteed all provinces 1993 levels of revenues in absolute terms in perpetuity.
Providing a share of the (increasing) revenues from the VAT with the provinces that generated the value added—mainly benefiting the richer ones.
Introducing a modern equalization framework enabled all provinces to provide similar services with similar effort. The version adopted in China was based on a simplified version of the Australian model and benefited the poorer provinces.
The most innovative component of the reform package was a revenue-return policy to return additional funds to where they were generated, that is, the better-connected provinces, but on a gradually decreasing basis. This measure was critical in concentrating resources toward production hubs, leveraging existing connectivity to generate investment, exports, and employment in the short term.
The VAT that China initially implemented applied only to manufacturing and imports, because of both administrative constraints and the need to leave at least one tax handle in provincial hands (namely the local business tax on services). Furthermore, the VAT was the investment type; that is, the VAT on capital purchases could not be offset against the VAT due on sales. This, again, was to meet revenue targets, and this formulation of the VAT was simpler for China’s nascent tax administration to enforce. Almost 15 years later, in the aftermath of the 2008 global economic crisis, there was pressure to reduce the cost of doing business and protect Chinese competitiveness. The investment-type VAT was converted to a consumption-type VAT, with VAT on capital purchases permitted to be offset against VAT on sales.
China conducted a further reform in 2015, similar to the reforms in Mexico in 2013/14,3, to integrate the subnational local business tax on services with the VAT. This reform aimed to reduce the costs of doing business, with the VAT on service inputs being offset against sales and exports. The VAT integration was also needed to create an integrated economic area—and to remove the borders around Shenzhen, a successful SEZ, to facilitate development of regional links needed for a high-tech zone. Business tax integration with the VAT was expected to reduce revenues, given the additional input tax offsets and refunds, yet revenues increased. As in Mexico, expanding the VAT to cover the full value chain made tax evasion more difficult for firms, thus improving compliance, including for income taxes.
Effects of the Chinese Tax Reform Package on Informality
In China, as in Mexico, major tax reforms designed to reduce the cost of doing business and revenue loss centralized the wide-area tax policy framework and integrated tax administration. The consequence has been a loss of simple subnational tax handles, constraining subnational access to credit and ability to finance basic services and essential investments. Cities have continued to rely on land value capture, particularly land sales, to finance investment.
Shenzhen is an example of a successful SEZ. Yet the borders around Shenzhen were removed after VAT reforms in 2015, because the central government began refunding VAT on exports regardless of whether the export was from an SEZ. At the same time, Shenzhen suspended the land sales model, because it became clear that younger families could no longer afford to live in the metropolitan area and were being forced into the informal sector.
Information generated by the VAT is also an important tool for monitoring investment and production of natural resources, including petroleum. Again, international financial institutions typically recommend that VAT should not apply on capital goods in the petroleum sector, especially because this is to be refunded promptly on exports and is not a source of additional revenues. This practice, however, means that complete information on operations of the petroleum sector is not collected. As demonstrated in the landmark Ribadu Committee Report (PRSTF 2012) on the Nigerian petroleum sector, revenue losses could be averted if complete records were available and the VAT and treasury single accounts would play a role.
The more appropriate mechanism, especially in countries with prevalent informality, may well be what is practiced in China—to apply VAT on capital and other inputs in the petroleum sector to be offset against VAT liability on outputs, or zero-rated when exports are verified. As recommended in the Ribadu report (PRSTF 2012), tax systems must also track the flow of funds, including between petroleum companies’ and governments’ financial information management systems.
Whereas distributional effects depend on combinations of taxes and social policies, a simple VAT design with minimal exemptions reduces the need for compensatory measures. For instance, as in Mexico, excluding unprocessed basic staples (wheat, rice, corn) was designed to protect the poor (as also argued by Ahmad and Stern  in the context of India and Pakistan). In the final analysis, the best “safety net” available in Mexico is a combination of the nondistortionary minimum pension (65 years and older), that did not affect incentives to participate in the labour market, together with sustained job creation, especially in the lagging regions of southern Mexico. Similar issues arise with job creation in the western regions of China, and the absence of subnational own-source revenues is a major constraint to the “rebalancing” sought in China, especially in response to the pandemic. Mexico will face similar constraints, and connectivity investments are not likely to be sufficient, even if necessary.
Effects of the Chinese Tax Reform Package on Sustainable Development
The success of the 1993/94 Chinese reform is striking. In addition to the strengthening of the fiscal space, more than 750 million people were taken out of poverty, 500 million moved from rural to urban areas, and 150 million moved to the historic as well as new coastal hubs (for example, Shenzhen). However, now Chinese cities face difficulty financing sustainable basic spending. Local government land sales have led to urban sprawl and generated off-budget financing vehicles that created rent-seeking opportunities (Wang, Wu, and Ye 2018). Sprawl, in turn, has accentuated spatial inequalities, with adverse effects on the formal employment of migrant households (as is typically seen around megacities in emerging market economies, including in Africa). Although smaller interior cities also engage in land sales, lower appreciation rates make them less attractive to investors than metropolitan areas.
National investment in connectivity infrastructure is necessary for developing less well-off regions, but as the Chinese example illustrates, it is not sufficient to ensure balanced development or convergence. Firms are unlikely to move to smaller cities that lack adequate infrastructure and agglomeration economies, including availability of workers with the requisite skills. Workers may still want to migrate to megacities, drawn to better services and higher-paying jobs. Thus, the pressure of migration to megacities continues, with the likelihood that informal settlements will grow even larger.
Beneficial Property Taxes for Subnational Service Provision
Property taxes have long been regarded as the main source of finance for local spending, particularly in the United States. High property taxes in the US context are synonymous with good public service delivery. In most US cities, property taxes are linked closely with property values that reflect the quality of public education and other basic services.
A robust property tax system is also the foundation for ratings that govern municipal bond issuance for infrastructure investments. Mobility, as a key feature of political competition models, is significant: taxpayers will move to jurisdictions that provide better education, even if the property tax rates are higher. However, for the US-style property tax system to work, property titles must be clearly delineated. Furthermore, municipalities must be capable of real-time adjustment of property values.
Such conditions, however, do not exist in many parts of the world. The property tax is thus moribund in much of Africa, Latin America, and Asia. Experiments in China, in Shanghai and Chongqing, to introduce versions of US-style property tax have not been successful, given the political difficulty of taxing fixed-income households in potentially expensive areas, as well as defining property titles and valuations.
In emerging market and developing economies, with few exceptions, the “standard” property tax has been dismal in raising revenues. In Africa, only Mauritius (briefly) and South Africa exceeded 1 percent of GDP, and the Latin American average is just 0.3 percent of GDP. An additional problem is that the standard property tax is typically based on the US ownership and valuation model that discriminates against informal sector people who seldom own properties that they occupy. Informal migrants typically cannot use ad hoc contracts as proof of residence to qualify for loans and credit for purchase of assets or improvement of property, or even to access public services.
Given difficulties in establishing ownership and making timely valuations, the United Kingdom abandoned the ownership and valuation model of recurrent taxation of housing in the early 1990s and moved toward a “banded system” linked to the cost of local services or benefits provided by local councils. However, ownership and valuation were retained for business premises, which tend to have a more robust market, and all property sales.
In Mexico, rate setting at the state level with administration at the local level has also generated severe disincentive effects. Local administrations have no accountability, and proximity to taxpayers generates opportunities for rent-seeking and favoring friends and relatives. The property tax is largely dysfunctional, with collections below the dismal Latin American and Caribbean average of 0.3 percent of GDP (Ahmad, Brosio, and Pöschl 2015). In India, although the tax has been in force since the late 1870s, collections are lower than in Mexico. In both countries, this tax just does not work very well (Rao 2012). Like in the United Kingdom, reforming a long-standing tax requires strong leadership to initiate change—or a crisis like the pandemic.
China experimented in Shanghai and Chongqing with variants of the US-type recurrent property tax, but neither city was successful. The problem is not only a lack of information on who owns which property or what the market valuations might be, but also the political economy conundrum of fixed- and low-income households living in expensive neighborhoods.
Local governments in China rely mainly on shared revenues and transfers, as well as land sales in some cases, and there has been telescoping budgetary pressures at the lowest levels of government as additional functions are devolved. And local government borrowing was not permitted, except through local government financing vehicles. As mentioned above, land sales typically accrue to off-budget accounts of local government financing vehicles, are not particularly transparent, and are open to rent-seeking, often circumventing national constraints on debt issuance. The incentives from increased subnational budget pressures and poor information on the local government financing vehicles led to a huge increase in unregulated subnational debt, particularly after the 2008–10 global financial crisis, when local governments played a key role in countercyclical policies on behalf of the central government.
China introduced a municipal bond system in 2015 to reduce the growing problem of subnational debt. However, because there are no effective own-source revenues, the measure has not succeeded in reducing subnational risks. To be sustainable, a municipal bond system requires effective own-source revenue that allows a local government to set rates at the margin, even if a band is set by the national legislature. A band for some property sales taxes has already been centrally legislated: provinces choose a range within the band, then cities pick a rate. This mechanism could serve as own-source revenue if it were applied to a more substantive base, such as recurrent beneficial property taxation—even if, as in China, all administration is centralized.
A Beneficial Property Tax Linked to Public Services
An attractive alternative to the US-type property tax model is to bypass the valuation system altogether and link property-based tax to area and location, as well as the cost of public services. Such a Marshallian “benefit-tax” proposal can help governments overcome political resistance and link taxes paid to services provided (Ahmad, Brosio, and Pöschl 2015). The ownership titling difficulties can also be finessed, because the benefit tax can be based on occupancy rather than ownership (unless the property is vacant).4 Linking property taxes to public services is also current practice in the United Kingdom. Former Prime Minister Margaret Thatcher abandoned the traditional property tax because of difficulties in establishing ownership and keeping valuation changes current.5 Valuation requirements are no longer binding in a system that links property use to the cost of local services. However, an accurate map of properties is needed.
Satellite imagery can prove a useful tool and is now highly developed and readily available in most parts of the world. Satellite mapping of properties cannot be easily evaded, so governments can bypass the corruption that takes place in appraising property areas and structures. Another new area for research, being examined even in sub-Saharan Africa (see Ahmad, Brosio and Gerbrandy 2018), is the use of blockchain technology to register property transactions. With extensive property records, governments will be better able to detect evasion of both property and income taxes. Blockchain technology also has useful applications in monitoring forestry and natural resources, as well as stopping illegal logging and mining.
A close link remains, however, between own-source taxes, transfers, and expenditures. This is partly related to the resource envelope, but also the incentives local officials face. Local officials need own-source revenue handles if they are to be held accountable for local spending that comes with hard budget constraints. If municipalities are able to access deficit-filling transfers, the positive incentive effects of the proposed tax system will not materialize. The transfers generate perverse incentives, as local governments are loath to improve revenue collections for fear of losing reliable revenue that carries no local political costs. This perpetuates the culture of “fiscal dentistry”—transfers to fill deficits—that was once pervasive in India at all levels (Rao 2002).
In an assessment for China based on a six-city sample, Ahmad and others (2020) show that a simple area- and location-based tax on occupancy linked to basic services (including basic education) or benefits (including for informal sector workers) can raise revenues quickly (with a target of 2 percent of city GDP), significantly improve income distribution (demonstrated by major reductions in the Atkinson inequality measure, appropriately weighted), and anchor the system of municipal bonds, thus reducing financial risks. Given the location basis for setting tax rates, Guangzhou, the most economically advanced city, generates a much higher rate per square meter than the smaller and less well-off cities. This higher rate also sends taxpayers appropriate signals about the relative costs of migrating to the congested metropolitan areas along China’s eastern coast.
Yet, for a beneficial property tax system to work most effectively, the cost of public services should include a provision for informal sector workers, and the revenue base should be standardized according to an equalization framework that provides similar levels of services at similar levels of tax effort. As with any tax and transfer reform, the political economy aspects are critical. The reform must not create incentives that increase workers’ migration into the informal sector in metropolitan areas. The benefit-tax agenda consequently needs to be developed with the aim of creating smaller employment hubs—or “poles of development” as proposed in Senegal (World Bank 2016).
An occupancy-based property tax would also enable informal households to more easily access credit and essential public services, including preventive health care, strengthening the ability of urban areas to withstand future pandemics. In providing a basis for informal sector households to integrate with the established citizenry, establishment of an occupancy-based property tax is crucial for emerging market and developing economies.
Kireyev and Mansoor (2016), in keeping with the revenue potential estimates from the World Bank, propose a target of 2 percent of GDP from a property tax in Senegal. This is probably feasible in the medium term, but as Ahmad, Brosio, and Gerbrandy (2018) argued for Senegal and Tanzania, a target of 1 percent of GDP should be easier to meet at least initially. Administration of the tax could be facilitated with a new streamlined tax administration structure for the property tax based on satellite imagery and blockchain technology. However, linking the property tax with basic public services is critical in generating political acceptability.
Tax-Benefit Links at Different Levels of Government
A considered proposal for the property tax to finance local operations will likely include several components:
1. The area- and location-based tax on housing would not preclude the value-based approach for other types of property, such as for commercial enterprises.
2. The value-based property tax should apply to commercial properties, as the market is typically vigorous and the political economy constraint of low-income households in expensive neighborhoods does not apply.
3. All property sales should be subject to the value-based transfer tax, or stamp duty, as in China and the United Kingdom.
4. The area-based tax would be based on occupancy. An occupancy-based property tax would provide migrant households and informal sector workers with benefits, enabling them to apply for loans and access basic services for which proof of residence is required. Area-based and value-based property taxes, thus, should not be seen as mutually exclusive.
5. The beneficial aspect of the tax is critical and needs to be carefully calibrated. Providing benefits solely to poorer migrants might not be politically acceptable to the broad population of taxpayers, even if it reduces inequality. As seen in the China simulations, a wider benefit for all households would be inequality reducing and likely to reduce political resistance.
An occupancy-based beneficial property tax of the sort described would be attractive to all residents, including migrants and informal sector workers. By definition, targeting city revenues about 1–2 percent of local GDP would generate a significantly higher tax per square meter in richer cities, such as Dakar or Mexico City, than in smaller interior cities. Higher taxes would send a market-based signal to workers and temper incentives to migrate to already congested and polluted metropolitan areas.
Cities need a proper system of local own-source revenues before they are able to sustainably borrow for investment without adversely affecting the interests of migrants and informal sector workers. It is important to align incentives for a clean environment with employment generation in clean, compact, and connected hubs.
It is also important to recognize that public-private partnerships and municipal bond systems generate local liabilities, and repayment schedules need to be linked to own-source revenue generation. Otherwise, liabilities might be hidden until there is a crisis, and accentuate risks, as during the 2008–10 financial crisis in Europe (Ahmad, Bordignon, and Brosio 2016).
Two dimensions of informality are addressed in this chapter: (1) firms “cheating” to evade taxation on transactions, wages, and profits; and (2) workers engaging in often invisible activities, not paying taxes but also excluded from most public services and benefits.
The first type of informality is greatly facilitated by the second—that is, people willing to work without formal contracts or benefits generate rents for firms, but at a cost of greater inefficiency and loss of public revenues. The second, more typical sort of informality, while generating employment for migrants, also generates inefficiencies, low standards of living, and significant inequalities. Informal workers and their families are excluded from basic services and are vulnerable to shocks and disease, including pandemics such as COVID-19.
The focus on national tax instruments is driven by the need to raise revenues to finance basic general government spending needs and create a unified economic space while reducing the cost of doing business and generating information. The political economy context of reform involves an assessment of gainers and losers and entails a link with multilevel finance, ensuring that no subnational jurisdiction or significant group suffers net losses. Subnational tax instruments, closely linked to the delivery of basic services and SDGs, form a beneficial nexus that also helps address traditional informality.
There is increasing evidence that inappropriately designed policy prescriptions result in what Levy (2008) termed “good intentions, bad outcomes.” These policies are often borrowed from the most advanced economies to address income redistribution objectives. Yet exemptions and multiple rates in the VAT and complex means-tested benefits, including conditional cash transfers, may not result in the desired objectives while jeopardizing revenue requirements for the SDGs. Such “bad outcomes” were seen in Mexico and partially redressed in the 2013 reforms, as well as in countries like Indonesia and Pakistan.
A key element in generating revenues and also reducing cheating is information on the full value chain from a “clean” VAT without exemptions that affect business-to-business transactions. A clean VAT also reduces the cost of doing business and helps create a level playing field for investments and exports. This was the objective of the 2013 Mexico reforms, with the integration of REPECOS into the VAT (reversing a recommendation from international financial institutions to focus on large taxpayers) and the 2015 China reforms, when the local business tax on services was folded into the VAT on goods administered by the central government. In both countries, revenues increased significantly as the ability to evade income taxes decreased. This tax-on-tax interaction (the VAT tends to be proportional at best, or mildly progressive if basic food items are excluded) is indeed a better mechanism to raise income tax revenues than targeting nonwage sources of personal income that tend to be difficult to track.
The main drawback to centralization of broad tax instruments is that it leaves subnational governments with relatively few tax handles. This, in turn, impedes progress on the SDGs, which could most effectively be met in clean, compact, and connected cities and through local action. A local tax system should not only finance key components of the SDGs but also generate incentives and market-based signals to slow the growth of informal settlements in the urban periphery.
In China, centralization of broad tax instruments has reinforced land value capture methods, such as land sales, to finance local priorities (Wang, Wu, and Ye 2018). Despite some success, land value capture has resulted in growing urban sprawl with the development of megacities, loss of prime agricultural land, and encouragement of off-budget “public” activities of dubious benefit. The metropolises along the eastern coast of China, as with Dakar, Jakarta, Karachi, and Mumbai, continue to draw migrants, often in informal settlements around the formal city boundaries. Local governments relying on shared revenues in the face of the COVID-19 pandemic have faced additional constraints as central governments have reduced national tax rates to address trade and health care shocks— this is true in all countries. In China, local governments have tried to resume land sales.
Property taxes have been suggested as the most appropriate subnational tax handle, but the traditional ownership valuation model has not worked in most emerging market economies (including China and Mexico), generating almost negligible revenues at best. The ownership basis does not suit informal sector migrants, who would gladly pay a tax (Ahmad, Brosio, Gerbrandy 2018) to show proof of residence, to enable their children to go to school, to visit health clinics, or to qualify for business loans. But the tax-benefit linkage is crucial.
Ongoing research in China and Mexico6 has simulated the effects of a beneficial property tax based on a simplification of the current UK model. Occupants, not owners (unless the property is empty), would pay a tax per square meter, and this would be higher in the richer localities. Given the minimal information requirements, a beneficial property tax could be adopted quickly. Such a tax would improve access to public services and credit, which many informal households would welcome. The similarities between policy measures relevant for different political systems, as in China and Mexico, highlight the importance of own-source revenues and tax-benefit linkages.
Local beneficial property tax options and links with basic health care and education are critical in the agenda to recover from the global pandemic, to address sustainable employment and informality. This would supplement a national tax agenda that addresses incentives to cheat while reducing the cost of doing business and creating a unified economic space for sustainable growth.
Ahmad, E. 2015. “Multilevel Fiscal Institutions and Mechanisms for Reducing Tax Cheating: The Case of Mexico.” In Institutions of Intergovernmental Fiscal Relations, edited by J.Kim andH.Blöchliger. Paris: Organisation for Economic Co-operation and Development.
Ahmad, E. 2015. “Multilevel Fiscal Institutions and Mechanisms for Reducing Tax Cheating: The Case of Mexico.” In Institutions of Intergovernmental Fiscal Relations, edited by J.Kim andH.Blöchliger. Paris: Organisation for Economic Co-operation and Development.)| false
Ahmad, E. 2018a. “Political Economy of Tax Reforms: Improving the Investment Climate, Addressing Inequality, and Stopping the Cheating.” G24 Background Paper, Intergovernmental Group of Twenty-Four, Washington, DC.
Ahmad, E. 2018a. “Political Economy of Tax Reforms: Improving the Investment Climate, Addressing Inequality, and Stopping the Cheating.” G24 Background Paper, Intergovernmental Group of Twenty-Four, Washington, DC.)| false
Ahmad, E., andA.Mohammed. 2018. “Pakistan, the United States and the Bretton Woods Institutions: a continuing Great Game?” In Routledge Handbook of Contemporary Pakistan, edited by A.Pande. Routledge, New York.
Ahmad, E., andA.Mohammed. 2018. “Pakistan, the United States and the Bretton Woods Institutions: a continuing Great Game?” In Routledge Handbook of Contemporary Pakistan, edited by A.Pande. Routledge, New York.)| false
Ahmad, E., andH.Viscarra. 2020. “Distribution and Revenue Effects of a ’Beneficial Property Tax’ in Mexico.” LSE/CUT Program on Financing Sustainable Urban Transitions in China and Mexico, London.
Ahmad, E., andH.Viscarra. 2020. “Distribution and Revenue Effects of a ’Beneficial Property Tax’ in Mexico.” LSE/CUT Program on Financing Sustainable Urban Transitions in China and Mexico, London.)| false
Ahmad, E.,G.Brosio, andJ.Gerbrandy. 2018. Property Taxation: Economic Features, Revenue Potential, and Administrative Issues in a Development Context. Luxembourg: Publications Office of the European Union.
Ahmad, E.,G.Brosio, andJ.Gerbrandy. 2018. Property Taxation: Economic Features, Revenue Potential, and Administrative Issues in a Development Context. Luxembourg: Publications Office of the European Union.)| false
Ahmad, E.,G.Brosio, andC.Pöschl. 2015. “Local Property Taxes and Benefits in Developing Countries: Overcoming Political Resistance?” In Handbook of Multilevel Finance, edited by E.Ahmad andG.Brosio. Cheltenham, UK: Elgar.
Ahmad, E.,G.Brosio, andC.Pöschl. 2015. “Local Property Taxes and Benefits in Developing Countries: Overcoming Political Resistance?” In Handbook of Multilevel Finance, edited by E.Ahmad andG.Brosio. Cheltenham, UK: Elgar.)| false
Ahmad, E.,M.Niu,L.Wang, andM.Wang. 2020. “Financing Inclusive Urban Transition in China: The Role of a ’Beneficial’ Property Tax.” LSE/CUT Program on Financing Sustainable Urban Transitions in China and Mexico, London.
Ahmad, E.,M.Niu,L.Wang, andM.Wang. 2020. “Financing Inclusive Urban Transition in China: The Role of a ’Beneficial’ Property Tax.” LSE/CUT Program on Financing Sustainable Urban Transitions in China and Mexico, London.)| false
Fuentes Castro, H. J.,A. Z.,Carrillo, S. B.Cortés, andA. B.del Rivero. 2011. Estudio de Evasión Fiscal en el Régimen de Pequeños Contribuyentes. Instituto Tecnológico y de Estudios Superiores de Monterrey.
Fuentes Castro, H. J.,A. Z.,Carrillo, S. B.Cortés, andA. B.del Rivero. 2011. Estudio de Evasión Fiscal en el Régimen de Pequeños Contribuyentes. Instituto Tecnológico y de Estudios Superiores de Monterrey.)| false
Gao, Q. 1995. “Problems in China’s Intergovernmental Fiscal Relations, Tax-Sharing System, and Future Reform.” In Reforming China’s Public Finances, edited by E.Ahmad,Q.Gao, andV.Tanzi. Washington, DC: International Monetary Fund.
Gao, Q. 1995. “Problems in China’s Intergovernmental Fiscal Relations, Tax-Sharing System, and Future Reform.” In Reforming China’s Public Finances, edited by E.Ahmad,Q.Gao, andV.Tanzi. Washington, DC: International Monetary Fund.)| false
International Monetary Fund (IMF), Organisation for Economic Co-operation and Development (OECD), United Nations (UN), and World Bank. 2015. “Options for Low Income Countries’ Effective and Efficient Use of Tax Incentives for Investment.” Report to the G20 Development Working Group, International Monetary Fund, Washington, DC.
International Monetary Fund (IMF), Organisation for Economic Co-operation and Development (OECD), United Nations (UN), and World Bank. 2015. “Options for Low Income Countries’ Effective and Efficient Use of Tax Incentives for Investment.” Report to the G20 Development Working Group, International Monetary Fund, Washington, DC.)| false
Wang, W.,A.Wu, andF.Ye. 2018. “Land Use Reforms: Towards Sustainable Development in China.” In Fiscal Underpinnings for Sustainable Development in China: Rebalancing in Guangdong, edited by E.Ahmad,M.Niu, andK.Xiao. Singapore: Springer.
Wang, W.,A.Wu, andF.Ye. 2018. “Land Use Reforms: Towards Sustainable Development in China.” In Fiscal Underpinnings for Sustainable Development in China: Rebalancing in Guangdong, edited by E.Ahmad,M.Niu, andK.Xiao. Singapore: Springer.)| false
Oportunidades had a negative effect on incentives to participate in the labor market, generated no reduction in poverty in the poorest state, and had been open to diversion of funds in key states (as seen during 2012). Oportunidades was replaced in 2014 by Prospera, which focused on training and supporting small business to encourage participation in the labor market. Prospera was abolished by the administration of President Obrador in 2019.
The United Kingdom raised more with the property tax (4.0 percent of GDP) than the United States with the traditional property tax (about 3.5 percent of GDP). See the IMF’s (2014) Government Finance Statistics Yearbook 2013.
This chapter draws on an article prepared for the Group of Twenty-Four, “Political Economy of Tax Reforms,” which was presented at the IMF/World Bank 2019 Spring Meetings (Ahmad 2018a), as well as at an ongoing, joint London School of Economics and Coalition for Urban Transitions research program on financing sustainable development in China and Mexico (Ahmad and others 2020).