4. Reforming recurrent taxes on immovable property

Salience improves accountability and transparency at the cost of popular resistance

Recurrent taxes on immovable property are very salient and transparent, unlike most other taxes. The most relevant taxes in terms of revenue-raising capacity are levied on income, consumption or payroll (i.e. mostly as social security contributions). All these taxes are, in most cases, collected automatically through payments and/or withheld at source, which requires the taxpayer to take no additional action to pay them. In contrast, in most cases taxpayers need to take action to pay their recurrent property taxes. Furthermore, taxes levied on income, consumption and payroll are paid “continuously” throughout a fiscal year whereas the recurrent taxes on immovable property are often paid in a (a couple of) relatively larger lump-sum payment(s).

On one hand this salience and transparency can improve efficiency and government accountability but, on the other hand, they create fierce opposition to tax hikes. In principle it is good for taxpayers to be aware of their tax payments so they can demand an appropriate supply of public service. Since recurrent taxes on immovable property are usually accrued to local governments and since public local services tend to be more visible than public services provided by upper levels of government, it is possible to directly link payments to a visible improvement in the supply of public services. There is evidence that when taxpayers can visualise the benefits that they receive from paying taxes, they are more willing to pay them (Slack and Bird, 2014[3]).

Offering the option for taxpayers to pay the tax in several instalments, withholding the tax at source and coupling the tax reform with improvements in local services are among the most common solutions to overcome the problem related to property tax salience. Although payment in instalments neither reduces the tax burden nor the transparency, the payment in multiple instalments reduces the value of each payment, which reduces its salience. Withholding the tax at source or coupling the tax payment with other payments can also help in reducing its salience but in this case at the cost of a reduction in transparency. For instance, property taxes can be paid in the same bill as mortgages and utilities. In some cases, it can also be withheld directly from income (e.g. see Box 4.1 in the Irish case). Lastly, coupling the tax reform with improvements in local services follows the benefit principle of taxation in which the tax payment serves as the payment of a public service, improving accountability. Nevertheless, it is also the most difficult to implement because it depends on activities beyond the scope of tax policy and because service improvements often require some time to become apparent whereas tax levies are immediate.1

Liquidity: the imperfect link between property values and income can create liquidity problems

Unlike other taxes, taxes levied on properties are often based on a stock rather than a flow. Although the value of properties is indeed correlated with the income of taxpayers, these two numbers are not perfectly associated. Thus, the tax obligation might be particularly large as a proportion of taxpayers’ income for asset-rich income-poor households. Usually pensioners are in that group, which might be one of the reasons why they tend to oppose recurrent property tax reforms.

Not only pensioners suffer this problem though – other taxpayers that lose either temporarily or permanently their source of income may also have liquidity problems to pay their property tax obligations. In 2020 this problem occurred in many countries due the COVID-19 outbreak. The outbreak reduced the income of some households but not the value of their properties (at least in the short term), creating a temporary mismatch between their property tax obligations and their financial situation. To cope with this problem, some countries decided to defer or even exempt a portion of property tax payments, such as Chile and Italy (OECD, 2020[6]).

There are multiple ways to solve this problem such as by granting: deferrals for seniors, tax exemptions or rebates based on income and more payment options. The case for deferrals is strong since they do not change the tax obligation, they only match the tax obligation payment with the period in which the taxpayer is expected to have a better ability to pay for it. Nevertheless, such schemes are not particularly popular among taxpayers and are thus not popular politically, given that the elderly, which compose the group of asset-rich income poor households, typically wish to leave their property for their heirs without substantial amounts of obligations (Slack and Bird, 2014[3]). As a result, taxpayers might be disinclined to take advantage of tax deferral programmes, which don’t solve the problem of the mismatch between property tax liabilities and current income. For these reasons, which could be the most economically efficient solution has limited power to weaken the resistance of the elderly against recurrent taxes on immovable property. Countries usually handle this problem through tax reliefs programmes. Some focus directly on pensioners (e.g. in Australia, Chile, Estonia, the Slovak Republic and some U.S states) while others on low-income households (e.g. in Estonia, Italy, Japan, Portugal, the Slovak Republic and the United States).2

Perceived regressivity of recurrent taxes on immovable property can be minimised with tax rate progressivity, tax reliefs to low-income households and bundling strategies

Recurrent taxes on immovable property are normally perceived to be regressive. As discussed in the second chapter, the empirical research on tax incidence suggests that the property tax can indeed be anything from progressive to regressive because the redistributive effect of property taxation hinges on various factors, such as the distribution of home ownership and tax design. In addition, recurrent taxes on immovable property are considered to have a smaller impact on the income distribution than other important taxes, notably income taxes. As a result, recurrent property taxation can be seen as regressive even if their impacts are not necessarily so.

Most countries aim at making their recurrent taxes on immovable property more progressive/less regressive by either levying progressive tax rates or by granting tax reliefs or exemptions to low-income households (to recap see Figures 2.4 and 2.5) It is worth noting that, in the latter case, the problem of asset rich and income poor households is also alleviated. Progressive rates can indeed increase the progressivity of a property tax system but, in some situations, it can aggravate the problem of asset rich income poor households.

Another potential solution is bundling a reform of recurrent taxes on immovable property with other taxes (e.g. income tax, wealth tax and inheritance tax) in a manner that the overall impact of the reform improves the income distribution even if the impact of the reform of recurrent property taxes alone would not. Property tax reforms can also be bundled with policies outside the realm of taxation – for instance, the reform can be bundled with an income-transfer programme targeted to the poor and, thus, although the property tax will not be progressive, the overall fiscal system will be (Kitchen, 2012[7]) or with provision of other public services, such as education (Ahmad, 2021[8]). Cash transfers to low-income households tend to be among the best tools to reduce income inequality (see Chapter 2).

In China’s case, recurrent taxes on the ownership of properties tend to be more progressive than for other countries. As discussed in the second chapter, since real estate represents most of the wealth of Chinese households, recurrent taxes on immovable property can have a greater effect in reducing income inequalities (OECD, 2019[9]). This prominence of real estate in household wealth can be highlighted in the discussions about the reform, potentially raising its popular appeal. Linking the reform to the provision of public services can have, though, greater redistributive effects.

The presumptive nature of recurrent taxes on immovable property can lead to unfair assessments or perception of unfairness – both of which can be mitigated with a good tax administration

Property values are a relative concept – different buyers may value the same property differently and, thus, rarely can one estimate a value for a property that is considered fair by everyone. In recurrent taxes on immovable property, property values have to be estimated to serve as a basis for the tax obligation. This problem does not occur, for instance, in transaction taxes in which the taxable value is the value agreed between the buyer and seller. The necessity of having an estimation of the tax basis can lead to the situation in which taxpayers believe that the value of their properties is not being correctly assessed. can contribute to the tax being perceived as unfair, which may create problems in property tax reforms (see Box 4.2 on Latvia’s case).

Multiple tools can be put to use to reduce the perceived unfairness in property assessments – for instance, taxpayer education, involvement of the taxpayer in the valuation process, accessible appeal process and conservative valuations. All these solutions involve a continuous effort from the tax administration. A transparent administrative process that timely makes information available to taxpayers is crucial for the tax system to be perceived as fair. For that purpose, tax bill notifications may serve as a communication channel, in addition to other channels such as television, newspapers and posters advertisements. Communication can focus not only on explaining the process of the tax collection but also on explaining the role of the property tax in funding public services (Box 3.5, in the previous chapter, covered this issue in more detail). Communication should also occur both ways, by giving room for the taxpayer to ask questions and query the tax department, potentially leading to changes in assessments if the taxpayer has a strong point.

Lastly, conservative valuations3 (i.e. that aim at estimating a value slightly below properties’ values) might also significantly reduce appeals since when taxpayers believe that the assessed value is below market values they are less likely to appeal. Although this practice can reduce tax revenues,4 it can also increase compliance and reduce the appeal rate, both of which are costly to the tax administration.

Erratic volatility: tax obligations can vary abruptly when cadastral values are rarely updated, creating resistance against property revaluations

Although property values tend to change over time, unless they are updated in the fiscal cadastre, tax obligations are not going to reflect properties’ value. The longer the period in between reassessments, the greater the mismatch. And the greater the mismatch, the greater the tax hike when properties are reassessed. Abrupt tax hikes are unpopular. The end result is a reinforcing feedback loop, in which the more cadastral values are outdated, the more likely for their update to be unpopular, reducing the likelihood of a reform’s approval and implementation, perpetuating distortions.

Tax hikes caused by property revaluations are also not uniform: shifts in taxes depend on the movement in housing prices that can differ markedly by region or even neighbourhood. Thus, even if the average house has not appreciated much, the tax hike can be significant for some taxpayers. And, again, since income and property values are not perfectly linked, this increase in tax obligation can lead to liquidity problems and face fierce opposition, as was the case in many reforms (Slack and Bird, 2014[3]).

In principle, the best solution for this problem is to keep cadastral values updated. This requires frequent reassessments that can be costly. Indexing between reassessments can also help but if used alone may lead to distortions since indexing, by definition, uses an index that represents an overall price change and house prices’ movements are asymmetrical by region, infrastructure, among others. Therefore, indexing can be used as a palliative solution but, if used alone, will lead to horizontal inequities and, in the long run, a tax hike for houses that experienced a greater increase in their market values in comparison to their cadastral values may still lead to abrupt tax hikes. Phased increases can also help but, again, should be very well structured to not create distortions (see Box 4.3 on Denmark’s case).

On the one hand, inelasticity with regard to the cycle can create a mismatch between tax revenues and government expenditures but, on the other hand, can serve as a cushion in times of crisis

The fact that property tax bases are relatively inelastic with respect to economic activity can create long-term mismatches between property tax revenues and government expenditures. Belinga et al. (2014[11]) estimated the buoyancy of different types of taxes in OECD countries and concluded that, with regard to property taxes, their short-run buoyancy is the only one that is not statistically significant, and that their long-run buoyancy is the lowest among other taxes (i.e. personal income tax, social security contributions, corporate income tax, and good and service tax). This can be explained by three factors: 1) the imperfect link between property values and economic activity; 2) the mismatch between house prices and cadastral values; and 3) local policy makers in some countries (more notably in the United States) often reduce statutory tax rates when properties’ market values are raising in order to not increase the tax burden on dwellers while maintaining their property tax revenues.

Despite these factors, policy makers and tax administrators can make property taxes more buoyant through different means. First, regarding the imperfect link between property tax bases and economic activity, policy makers can select a more “buoyant tax base” – property taxes applied to capital and rental values tend to be more buoyant than area-based ones, since the latter has no significant link with the economic activity while it is expected that a booming economy would increase house prices and rents (Almy, 2014[12]). Second, the mismatch between house prices and cadastral values can be minimised if cadastral values are indexed in-between reassessments and/or properties value are reassessed frequently (see Figure 3.3, in the previous chapter, to see examples of countries that implemented such measures). Third, tax rates should, ideally, not be reduced in times of house price booms.

This last recommendation is disputable and it is worth digging into it further since inelasticity with regard to the economic cycle reduces the impact of crises on property tax revenues. As a result, all other things held equal, the higher the share of recurrent taxes on immovable property revenues in government revenues, the less they are impacted by economic crises – see OECD (2020[13]). Given that local governments tend to have less access to financial markets than central governments, which raises their vulnerability to liquidity crisis, this feature of recurrent taxes on immovable property is especially valuable for them.

In that light, Belinga et al. (2014[11]) suggested that municipalities might aim at keeping property tax revenues stable, regardless of what happens to GDP (and house prices). Norregaard (2013[14]) showed that even in periods of very rapid property price appreciation, property tax buoyancy still remains low. Lutz (2008[15]) performed empirical research for US states and concluded that the elasticity of property tax revenues with respect to home prices equals 0.4, indicating that statutory tax rates are probably adjusted downwards when house prices rise, reducing the tax buoyancy and stabilising property tax revenues. Norregaard (2013[14]) also found that SNGs in the United States tend to compensate for revenues losses in times of crisis by increasing property tax rates. In this situation, property tax rates tend to behave counter-cyclically – they increase in times of crisis and decrease during economic upswings.

Despite these benefits that stem from having a stable revenue source, it cannot be overstated that the main goal of a tax is to raise revenues for governments. Taxation is the main means through which governments can fund public services. Thus, the long-run buoyancy of property taxes should, ideally, not be very small. Very small long-run buoyancy would reduce the capacity of property taxes to raise revenues and, thus, would harm the funding of essential public services. In case other taxes are adjusted upwards to offset this small buoyancy of property taxes, there will be a neutral shift towards these other types of taxes, changing the properties of the tax system, potentially in an undesired manner since this change was driven only by the necessity to raise more revenues. The use of a more buoyant tax base and frequent property revaluation (possibly with the use of indexed cadastral values between revaluations if they are not frequent enough) are, thus, recommended.

The use of assessment limits, property tax capping, banding, classified tax rates and self-assessments can create distortions that, in the end, might cause more harm than good

Some countries employ solutions that do solve one of these specific problems but at the cost of creating others. Caution is advised when employing these palliative solutions. Among them there are the imposition of assessment limits, property tax capping, banding, classified tax rates, use of self-assessments for property assessments and the sole use of indexation for updating cadastral values. The reasons are provided below – based on Slack & Bird (2014[3]):

• Assessment limits (i.e. restrictions to how much properties’ cadastral values can increase) and property tax capping (i.e. restrictions to how much tax obligations can increase) both aim at cushioning property tax hikes and, thus, they also have a role in reducing the chances of causing liquidity problems to taxpayers when tax obligations increase. Nevertheless, these benefits come at the cost of horizontal inequities. By capping increases in property values, some properties are going to be undervalued in comparison to others (see Box 4.4 on the Californian case). More specifically, properties whose values are growing rapidly will be undervalued in comparison to properties whose values are either stable, growing slowly or decreasing. As a result, taxpayers that are having their assets appreciating the most are the ones who benefit the most from these restrictions, which may create distortions that reduce the progressivity of the tax.

• A similar outcome occurs when tax obligations are capped: taxpayers who should be taxed the most are those with higher property values and, thus, they benefit the most from tax capping. Moreover, capping taxes obligations, either directly or through limits on assessments, are also capping tax revenues. In the long run, limits to tax revenues can create mismatch between tax revenues and government expenditures. In addition, paradoxically, because such measures reduce the elasticity of property tax revenues, they likely increase the pressure for future property tax reforms (Slack and Bird, 2014[3]).

• Banding, which is used, for instance, in the United Kingdom and Ireland, can alleviate the impact of property value hikes and reduce administrative costs since banding systems are rather simple to understand and to administer. Nevertheless, banding creates horizontal inequities caused by the impossibility of valuing properties in such a discrete manner – the range of values (or values per square meters) of houses vary markedly, which is not captured by the small number of bands (e.g. eight values in the United Kingdom and nineteen in Ireland).5 This discrete nature of this system is particularly perverse for taxpayers located in a boundary of a band, leading to appeals and discontent.

• Self-assessments are useful for dealing with the presumptive nature of the property tax and also for reducing administrative costs but these benefits come at the cost of precision. Self-assessments should not be confused with consultation. The former regards the situation in which taxpayers provide the value of their property whereas the latter refers to the case in which taxpayers provide additional pieces of information that are used in the appraisal. Consultation is a good tool to gather data in a relatively cheap way while improving the credibility of the appraisals in the eyes of the taxpayers given that they participate in the valuation process. Self-assessments can also be useful when employed temporally, but when employed permanently and as the only means to gather properties’ information, they are likely to lead to inaccurate estimates of property values, with a tendency toward under-estimation6 (Slack and Bird, 2014[3]). In case governments are to audit self-assessments to handle these issues, they might significantly increase costs and create discontent, defeating the purpose of setting a self-assessment system.

• Classified tax rates (i.e. the situation in which multiple tax rates are applied to different “classes” of properties) can be used to increase progressivity in case tax rates vary with variables linked to income. When varying tax rates apply to different properties whose incidence rests on taxpayers in a manner that the higher the rate the higher taxpayers’ ability to pay, in theory, we have a progressive tax system. Nevertheless, classification of properties can significantly increase the complexity of the tax system and create room for arbitrary classifications (see Box 4.4 on the Ontario’s case, below, and Box 3.6, on the Israeli case, in the previous chapter). The higher the complexity of the tax system the lower its transparency and the higher its compliance costs, and the higher the room for governors/policy makers to classify properties and to define tax rates to each type of property, the more the tax system can be exploited to benefit some taxpayers due to political reasons, generating horizontal inequities.

• Indexation, as discussed throughout this section, can yield good results if used in between frequent reassessments. It does increase buoyancy and it also averages out strong tax hikes. Nevertheless, the sole use of indexation will create horizontal inequities in the long term since the effect captured by the index represents an average and not individual house prices. House prices’ movements vary markedly depending on many aspects of a property such as its region, neighbourhood, infrastructure, among others and, thus, cadastral values that were only updated though indexing will lead to unfair assessments in the long term.

Timing: economic upswings and booming housing markets can create an opportune moment for putting forward a property tax reform, yet reforms following downturns can also be successfully implemented

Kingdon (2014[20]) created a theory that explains the process through which a policy is designed and implemented – the Multiple Streams Framework. This framework outlines that a solution to a problem in the format of a policy is only implemented when three different streams come together at the same moment: the problem, policy and political streams. The problem stream refers to the public and media attention given to a specific problem. The policy stream refers to the existence of solutions to address this particular problem. The political stream regards the policy makers’ motivation to implement a project with such a solution. Therefore, according to this theory, a policy is only implemented when, at the same moment, the people/media are giving attention to a particular problem for which there are available solutions that can solve the issue and the policy makers have incentives to turn at least one proposed solution into an implemented policy. This moment when these three streams come together is called a policy window.

Policy windows are not necessarily created in a spontaneous manner Kingdon (2014[20]). In fact, “policy entrepreneurs” are actors whose role is to create policy windows so a policy can be successfully launched into the political arena. Tax reforms are no different. Tax reforms that are well-designed to solve the underlying issue and that are backed by the general public and policy makers have greater chances of approval. As a result, it is crucial to put the reform forward only when these elements are in favour of the tax reforms. For these reasons, reforms are prepared in advance and usually wait for a good timing to be put forward (Brys, 2011[1]). Besides, there are multiple aspects of a reform that depend on timing such as when the reform is discussed with policy makers, when it should be discussed with the general public, when it should come into force, among others. Regarding the latter (the implementation), it is worth highlighting that the implementation of a tax reform can have multiple phases, and the implementation of each phase also depends on good timing to be successful. Not rarely reforms are phased and there is no support for the latter incremental stages of a reform, leading to an incomplete reform (see Box 4.6 on the Korean case at the end of this section).

Some particularly relevant issues for timing a property tax reform are the states of public finances, economy and housing market. Experience shows that it might be easier to implement fundamental tax reform when a country is running budget surpluses that could absorb possible revenue losses or could be used to partly compensate the “losers” from the tax reform (Brys, 2011[1]). Regarding the economy, due to the salience of property taxes, introducing revenue raising property tax reforms in economic downturns can harm households whose income have shrunk and, thus, not able to pay the taxes. In other words, the reform would not only have less compliance but could also have an undesired pro-cyclical effect. Therefore, there are incentives for governments to engage in property tax reforms in good times.

For instance, Blöchliger and Vammalle (2012[2]), in a study on fiscal federalism and subnational government reforms covering ten OECD countries (Australia, Austria, Belgium, Canada, Denmark, Finland, Italy, Portugal, Spain and Switzerland), found evidence that a reform succeeds if implemented at times of economic upswings and healthy fiscal conditions. As the authors put it “while a few reforms were initiated when the economic situation was weak or driven by consolidation needs, implementation of virtually all reforms took place when central and – to a lesser extent – sub-central public finances were in good shape. A sound fiscal position allows the central government to compensate sub-central governments, while reform resistance tends to be stiffer and reform failure more likely without central government support.”

Concerning the housing market, evidence shows that recurrent taxes on immovable property can be used as a policy instrument for asset price stabilisation (Blöchliger et al., 2015[21]) and can also influence the share of housing investments in total investments through different tax treatment (Millar-Powell et al., 2022[22]). Rapid tax increases during a slowdown – or tax reductions during a boom – seem to make property taxation strongly destabilising (Blöchliger, 2015[4]). In that light, when the housing market is in a downturn, the potential contraction caused by revenue-raising property tax reforms can attenuate the contraction, which is undesirable. On the other hand, when the housing market is booming, a reduction in the growth pace can be desirable.

Despite this solid case for putting forward property tax reforms in economic upswings, downswings can also improve the political support for a tax reform through various channels (see Box 4.5 on the Italian case). Pro-growth reforms can gather more political and popular support in downswings since they can be employed to revert the situation. More specifically, recessions sometimes expose very clearly the weaknesses of the economy and, thus, the need for reforms, leading to a greater acceptance of the necessity of the reform from taxpayers and voters (Brys, 2011[1]).

The timing to start discussing with the general public and/or experts also varies greatly. One approach is to present the tax reform only when their main elements are already defined and there is little uncertainty about their impact. Combined with good communication, this strategy may increase the likelihood of a reform to resist political pressures due to the reform’s maturity. Another approach is to start the discussion earlier and let the learning occur during the discussions. This approach is especially useful in the presence of high levels of uncertainty about the number of winners and losers and the extent to which they are (positively or negatively) affected by the tax reform. That is because in this situation policy makers might become more careful in making a decision to engage in tax reform (Brys, 2011[1]). Therefore, anticipating discussions can lead to a long period of debate with the purpose of answering the most difficult questions and criticisms. However, after this long period the reform may reach a maturity, leading to a quick implementation.

Administrative problems can also affect a reform’s timing. In that regard, tax reforms in general have been most successful when they are least needed. Rosengard (2012[23]) explains that non-urgent reforms allow tax administrators to introduce new processes smoothly and, in that manner, the learning period for both taxpayers and administrators to familiarise themselves with the new system can last longer, enhancing the understanding of, and voluntary compliance with, the tax. Timing for taxpayers to familiarise with the tax is especially relevant for recurrent taxes on immovable property since taxpayers usually have to actively pay the taxes (in contrast to income or consumption taxes that are, in most cases, paid indirectly when receiving income or buying a product, respectively).

In light of this discussion, it can be stated that there seems to be no perfect time for putting forward a property tax reform. Despite that, we argue that introducing a recurrent levy on residential households in the near future can benefit China in many different ways. First, despite of COVID-19’s contractionary effects, China is not facing a significant economic downturn – on the contrary, the Chinese economy is recovering at a strong pace, which is particularly important for the reform to not have pro-cyclical effects. Second, house prices in China have continued to rise throughout the last decade. Although the upswing movement was stronger in the early 2000s, prices are still growing in most regions. Third, the vertical fiscal gap is widening in China (CDRF, 2020[24]), and, thus, a new source of income could potentially revert this trend. Although Chinese local governments may not have the necessary income to compensate losers during the first phases of the implementation, these compensations can be temporarily made, if necessary, as a part of an agreement with upper levels of government. There are no guarantees that this positive outlook will continue in the coming decades and, thus, it is opportune for China to introduce recurrent taxes on immovable property relatively soon.

Administrative hurdles: a collection-led strategy coupled with a focus on more valuable properties can help emerging economies to maximise tax revenues

Recurrent taxes on immovable property are among the taxes that depend the most on having an effective administration to translate tax laws and regulations into field realities (Slack and Bird, 2014[3]). In some cases, the scope of recurrent property tax reforms covers only administrative issues (e.g. in case of reforms focused on updating cadastral values). In other cases, the administration needs to be reformed to operationalise the design changes. Rarely a property tax reform does not include any administrative element.

Often administrative reforms involve costly changes that require time to mature, such as the development of digital systems and the training or hiring of human resources. In addition to these costs incurred by the tax administration, changes in tax administrations might also increase taxpayers’ compliance costs, at least during the period when they are adapting to the reform (e.g. learning to fill new forms, understanding the new elements of the tax obligation, among others). In order to reduce such costs, reformers can implement the reform in a more gradual way through the use of a transition phase or through an incremental implementation of the reform (Slack and Bird, 2014[3]). Both strategies can be used to improve the link between taxpayers’ and officials’ learning curve with the reform’s pace. When these strategies are employed, the reform is only fully implemented in the medium-term.

Especially for governments in developing economies that do not have the capacity to register, valuate and collect revenues from all types of properties in a short period of time, a particularly interesting gradual implementation strategy is to focus on properties that can raise more tax revenues (Rosengard, 2012[23]). Particularly in developing economies, the distribution of properties’ values tends to be highly skewed, with a relatively small number of properties comprising the bulk of properties’ values. As a result, narrowing a reform’s focus to these small number of high-end properties can maximise tax revenues in the short term. In addition, administrative simplifications can be made in order to maximise net tax revenues from low valuable properties. For instance, self-reporting and a simplified CAMA can be employed for cheaper properties while individual valuations are only employed for more expensive real estate. This strategy can aid developing countries to raise tax revenues in a quicker manner. These revenues can then be used to provide the necessary local services to raise popular support for the reform and to finance the costs of the rest of the reform.

In China’s case, administrative capacity is asymmetrically distributed across municipalities. For instance, Shenzhen already has an advanced property valuation system in place (for more on Shenzhen see Box 3.4). Shanghai and Chongqing already have some experience with recurrent property taxes on residential properties since they already have introduced this levy in a small scale due to a pilot programme. In this situation, the central government can allow local governments to implement the reform in different moments – the greater the administrative capacity, the earlier the levy can be introduced in a municipality. For municipalities that do not have the capacity to implement the reform, Liu (2019[27]) suggests that a financial incentive programme in the form of an inter-governmental transfer aimed at rewarding municipalities that make substantive efforts to roll out the new property tax can incentivise the introduction of the tax and help boost capacity. In addition, the implementation in each municipality can also be phased, with a stronger focus on high-end properties at first.

Another interesting option is to use a relatively simple tax base at the time of the introduction of the reform, when fiscal cadastres are not yet completed, and the administrative apparatus has not yet matured. Along these lines, Ahmad (2021[8]) suggests that China could introduce a recurrent tax on residential properties that uses area as the tax basis. Such a tax is relatively easy and cheap to implement, as it is not necessary to assess properties’ values and, if combined with a linkage between revenues and local expenditures, could still increase the overall redistributive effects of China’s fiscal system. This approach could be valuable for some regions and municipalities in which data on property transactions and characteristics are not available. By following this approach, the tax could be introduced sooner and, therefore, raise own revenues for SNGs sooner than waiting until property values are fully assessed. It would be important to clearly announce that the simplified approach is a transitory one, with the longer-term objective to use a value basis. Over time, then, depending on the maturity of the regional/local property tax administration and on the quality of the data collected on properties, the local/regional tax administration could move towards a value-based approach.

Transitional measures can be useful to alleviate short-term effects of a tax hike, nevertheless reformers need to design and implement these measures in a manner that the full implementation of the reform is secured in the long term

Tax reforms can change the status quo to which taxpayers have already adapted, leading to undesirable disruptive effects. In the case of a salient levy as recurrent property taxes, a sudden tax hike can catch households unprepared and may increase delinquency (Slack and Bird, 2014[3]). Furthermore, a sudden tax hike can also impact the real estate market in case tax obligations are capitalised into prices, potentially destabilising the state of the real estate market (Blöchliger, 2015[4]). As Slack and Bird (2014[3]) put it, “failure to allow adequately for transitional problems and to cushion burden shifts adequately can be a fatal defect”. In addition, property tax reforms may change the relative tax power across jurisdictions and, in case they are revenue neutral, can cause a reduction in revenues in some jurisdictions. Examples of (temporary) transitional measures include reduction in tax rates and the provision of tax benefits/reliefs such as tax credits, tax deferrals, among others (Box 4.3, presented in the previous section, gives interesting examples of transitional measures employed by Denmark).

Blöchliger and Vammalle (2012[2]) in an overarching study on fiscal federalism and subnational government reforms covering ten OECD countries found that transitional compensation is almost always necessary. The target of these measures can be stakeholders or jurisdictions. In the cases analysed by these authors, the central government usually funded these transitional costs and opt-out clauses were provided. For instance, the Swiss established transitional “cohesion funds” to ensure that SNGs would suffer no losses in financial terms in the next 28 year after the reform.

One option to cushion the impact of a reform on taxpayers is by keeping effective tax rates low until the tax administration can have a good coverage (Rosengard, 2012[23]). When fiscal cadastres are not yet complete, properties’ values were not assessed or the tax collection is not working properly in some specific regions, tax obligations across properties will shift greatly only due to administrative obstacles. In case this transitional period lasts long, these inequities may start to influence taxpayers’ behaviour, creating economic distortions. Keeping effective tax rates low may minimise this risk.

Although cushioning the impact is desirable to avoid disruptive effects, measures employed for that purpose are neither costless nor risk free. First, the longer the cushioning mechanism stays in place the longer it takes to move to the desirable tax system. Second, some planned changes might never be put into place and, in numerous cases, temporary solutions are perpetuated (see Box 4.6 in the Korean case at the end of this section). Thus, although transitional measures might be politically important for the approval of a reform, reformers need to design and implement these measures in a manner that the full implementation of the reform is secured so it can achieve its initial goals (Slack and Bird, 2014[3]). One way to increase the chances of a full implementation is to include in the original legislation the expiry date of the transition period or the date in which incremental changes will enter into force (i.e. sunset provision or clause). In that manner, groups that benefit from the transitional measures would have to work harder to maintain their special treatment since it would require a change in the reform’s plans, which were made transparent and were legitimised in the legislation (Brys, 2011[1]).

Two specific elements of the Chinese tax system are worth mentioning in the context of transitional measures. First, China already levies transaction taxes on immovable property and, thus, right after the introduction of recurrent taxes on residential property, taxpayers might feel that the new obligations are unfair since they just paid their transaction taxes. Second, Liu (2019[27]) pointed out that due to the very low carrying costs of housing investments in China, many households own multiple properties, and, in this case, the introduction of a recurrent property tax applied to these households could create a disruption. In order to alleviate this sense of unfairness and to give room for investment portfolio adjustment, transitional measures can be employed. Liu (2019[27]) and Brys et al. (2013[28]) suggested that authorities can mitigate these problems by phasing-in the property tax reform, granting a temporary high basic tax-free allowance for the recurrent taxes on immovable property and/or starting with a wide tax base and very low tax rate. In case the reform is bundled with a reduction of the levies on properties transactions, a tax credit can be given to taxpayers who recently paid transaction taxes so that the new recurrent levy on residential properties is not considered unfair to these taxpayers. Such a levy needs to be temporary and, ideally, phased-out gradually over a few years.7

Communication strategies for tax reforms: engaging a wide range of stakeholders, obtaining a consensus on a reform’s broad long-term goals and focusing the discussions on the systematic rather than individual impact of the reform

Communication, when employed effectively, can help sustain the necessary political and popular support to ensure that the reform will be introduced and implemented in full. Slack and Bird (2014[3]) suggest that both policy makers and taxpayers are more likely to buy into a reform if they have been consulted and feel that their issues have been properly heard. In this light, a consultation previous to the announcement and implementation of the reform can reduce popular opposition.

A communication strategy that engages multiple stakeholders such as business, unions, special interest groups, academics and the broader public can create a feeling of ownership and reduce resistance (Brys, 2011[1]). Blöchliger and Vammalle (2012[2]) highlighted that external expertise can be particularly useful since they can provide impartial and unbiased scrutiny, raising the credibility among the public. These authors showed how external expertise was fruitful in reforming elements of the fiscal policy in several OECD countries such as Finland, Australia and Canada. It is worth noting that the language spoken by these experts are often different from the language spoken by the average citizen and officials. These authors found evidence that creating a credible independent panel consisting of experts and professionals who know how to communicate with non-experts can help the interpretation of the reform package by the general public.

Blöchliger and Vammalle (2012[2]) emphasised that although consultation is clearly beneficial, too much consultation can inflame opposition. From the cases analysed by these authors, the most successful consultation strategy involved parsimony with numbers related to the short-run impacts of the reform and a focus on discussing a reform’s overall/long-term goals. In other words, the long-term objectives were the focus of the communication instead of short-term distributional impacts. These long-term benefits can be communicated to the general public through promotional slogans such as “better services” and “more autonomy”. It is worth noting that although it is beneficial to highlight long-term benefits, reforms issues should not remain hidden – successful reforms often emphasise the long-term benefits without concealing negative impacts (Blöchliger and Vammalle, 2012[2]).

In the same vein, Brys (2011[1]) highlighted the importance of obtaining a consensus on these broad long-term goals of a reform and explored how this can be achieved through a clear communication focused on long-term objectives and trade-offs involved. Accepting specific constraints to the reform might also facilitate the reform’s implementation since stakeholders are now aware of what is negotiable and what is not. In the absence of broad strategic consensus, the same author suggests that governments can reduce uncertainty and begin to guide debates by announcing aspirational tax reform’s goals before presenting specific proposals. This announcement can 1) bring support to the reform from groups that are benefited by the reform; and 2) create a reputational cost for the government to change the reform’s goals after an announcement.

Lastly on communication, Alt et al. (2010[29]) argue tax reforms should not be discussed in isolation. The general public often lacks understanding about the interplay between the tax being reformed and fiscal policy’s goals and, thus, can make wrong assessments about the impacts of the tax reform. Keeping a narrow focus on an individual tax hinders comparisons and makes it easier for opposing groups to attack the reform by showing its negative impacts without comparing them to the impacts of other available options. For instance, according to these authors, one of the reasons for the abolition of the estate tax in the United States (i.e. inheritance tax) was a narrow-focused communication, which helped the organisation of opposing groups that focused the debate only on the tax’s problems and not on its overall effect on the progressivity of the tax system and on its revenue-raising capacity that can raise funds for the provision of additional public services.

Bundling is commonly employed in recurrent property tax reforms as a way to overcome political resistance through compensation of losers and improving distributional effects

Bundling multiple reforms together is a good strategy to offset or alleviate negative impacts of individual reforms with the impact from other reforms. It is worth noting that bundling is not the same as transitional compensations as bundling may involve elements outside the original scope of the property taxation reform and in a permanent way. A bundling strategy can be designed in a manner to overcome political resistance and build popular support for the implementation of the reform. According to Olofsgård (2003[30]), bundling reforms may be especially useful in case there are many actors whose support is crucial – for instance, when multiple groups have veto power because each group should be satisfied in order for the reform to pass and, under these circumstances, bundling the reform with elements that satisfy each of these groups can secure a reform’s approval. Blöchliger and Vammalle (2012[2]), on the other hand, although they recognise that bundling is typically employed and often fruitful, it can lead to an excessive focus on satisfying stakeholders, detracting the reform from its efficiency-enhancing aspects.

Brys (2011[1]) explored two important benefits that come from bundling. First, it may make it easier to address distributional issues, especially when a tax is bundled with measures from the expenditure side that have substantial impact on the income distribution, such as cash transfers to the poor (Joumard, Pisu and Bloch, 2012[31]). Second, bundling tax reforms with improvements in service delivery can increase the overall acceptance of and compliance with the tax being reformed – see Simonsen & Robbins (2003[32]). For the latter strategy, it is particularly interesting to phase in tax changes over several years so the pace at which services can be improved is linked to the pace at which citizens are being taxed.

In OECD countries it is common for countries to bundle reforms in recurrent taxes on immovable property with other tax reforms and with changes in the inter-governmental fiscal framework.8 For instance, in 2005 Korea introduced a comprehensive real estate tax together with a reduction in property transaction taxes and with an introduction of a grant inversely related to property tax raising capacity (i.e. as a compensation for local revenue losses). Another example is Finland, which has, in less than 3 years, 1) reformed its property tax system; 2) increased the share of corporate income tax that is allocated to municipalities; and 3) discontinued its revenue equalisation transfer system (see Box 4.7 in the Finnish case for more details).

Leaseholds and collectively owned land can create problems for property tax reforms; nevertheless, these can be mitigated with bundling, communication and transitional strategies and with a clear distinction of the levies on land and buildings

When introducing recurrent taxes on immovable property in a country that has a leasehold system in place, taxpayers might oppose the reform on the grounds that they are being double taxed since they are paying both a lease fee and a recurrent property tax. Hong (2012[33]) raised interesting points, here highlighted, regarding the introduction of recurrent property taxes in such a context.

First, despite the convenience of combining all land related levies into one payment (i.e. lease and recurrent property taxes), taxpayers might confuse taxes with leasehold fees and, thus, these should be kept separate, so the taxpayer knows what he is paying for in each payment. The similarity of lessees and taxes payments without proper taxpayer education can create an expectation that the land is owned by the lessees. According to Hong (2012[33]) this has happened to some extent in Canberra/Australia and Israel, both of which have a leasehold system. In order to minimise this problem, lessees in Canberra must pay the entire leasehold value up front and the property tax afterwards on an annual basis (Neutze, 2003[34]). In addition to clearly separating the leaseholds fees and tax payments, taxpayers should be educated about property taxes and the leasehold system so they understand what the purposes of each charge are.

Second, when ownership of land and buildings are not the same (as in China, especially in the case of collectively-owned land), two different levies – one applied to buildings and other to land – can be created to tax these assets separately. These arrangements are not uncommon in OECD and partner countries.9 For instance, Hungary has a separate levy on land and improvements. Denmark and Slovenia have a separate land, building and immovable property tax. Greece, Ireland and the United Kingdom have a building tax plus an immovable property tax. Finally, Australia, Brazil, New Zealand, Poland and the Slovak Republic have a land tax plus an immovable property tax.

Third, as explored in the second chapter, tax liabilities can be capitalised into house prices10 and, in case of taxes applied to leaseholds, this capitalisation can hurt leasing fees. In the most extreme scenario, the revenue obtained from recurrent property taxes can be fully offset by a reduction in leasing fees. Of course, the timing of the capitalisation is not immediate and depends greatly on many factors such as the date of leaseholds’ renewals and the elasticity of supply and demand of properties. As a result, until the effect kicks in, it is possible for governments to use the revenues raised by the recurrent property taxes to improve the supply of public services and local infrastructure, among others. In that way, this improvement can also be capitalised into property prices, alleviating or potentially cancelling out the reduction in leasing fees. It also is worth noting that there is evidence that taxpayers are more inclined to pay property taxes when they rate the quality of public services highly (Simonsen and Robbins, 2003[32]).

Fourth, recurrent taxes on immovable property can be levied on the occupancy of the property – as suggested by Ahmad (2021[8]) – and, therefore, be interpreted as a tax on housing services or on capital or even as a fee for using land or local services. Although in essence the levy is applied to the same tax base, these different interpretations are particularly important in the context of leaseholds since some of these interpretations can minimise the impression of double taxation. Especially when property taxes are viewed as a fee charged to the occupant for using local services, this interpretation is implausible. That is because leasehold fees and local service fees clearly do not refer to the same charge. In that regard, one potential way to reduce the room for interpretation that recurrent taxes on immovable property lead to double taxation is to name the charge in a way that favours the interpretation of a fee for using local services. For instance, recurrent taxes on immovable property in the United Kingdom and in Australia are known as “council tax” and as “council rate”, respectively.

As a result of the third and fourth points, a bundling strategy that combines the introduction of the levy on properties with improvement of local services may not only offset potential reductions in leasing fees but also makes it harder to interpret the levy as a form of “double taxation”, reducing the resistance against the reform. In that context, Ahmad and Colenbrander (2020[35]) suggested that China could link property tax revenues with social benefits, such as free or subsidised transport for the elderly or support for basic education. These authors highlighted that this linkage could have many benefits such as: 1) improving the distributional impact of the tax; 2) enhancing accountability; 3) assuring a stream of funding for basic services; and 4) enabling access to private finance for investments. All these effects could increase the acceptability of the reform.

Fifth and lastly, the introduction of levies on leaseholds may be considered unfair to lessees who recently paid the entire leasehold fees, and thus, did not anticipate this future liability. In case leasing fees are reduced due to the introduction of the tax, this problem can become even more apparent. In that regard, transitional measures targeted at these taxpayers can be employed (the previous section covered this issue in more detail).

Recurrent property tax reforms can affect the distribution of revenue across levels of government (vertically) and within levels of government (horizontally)

One last issue that is worth discussing in the context of property tax reforms regards their impacts on different levels of government and inter-governmental fiscal relations. As recurrent taxes on immovable property typically are an important source of revenue for subnational (mainly local) governments, property tax reforms can significantly affect the distribution of revenues across levels of government (i.e. vertically) and within the same level of government (i.e. horizontally). Therefore, in order to alleviate negative impacts on certain levels of government and/or jurisdictions, it is common for governments to bundle reforms on recurrent property taxation with reforms on inter-governmental relations (see, for instance, the Korean case in Box 4.6, above, and Box 4.7 and Box 4.8 on the Finish and Italian cases).

Regarding vertical relations, when recurrent taxes on immovable property accrue to SNGs, raising their share in total taxation will tend to increase the share of subnational revenues in general government revenues. This shift in the distribution of revenues across levels of government as a “by-product” of a property tax reform can be undesirable. In this situation, this relative increase in subnational revenues can be offset by a reduction in inter-governmental grants from upper levels of government. In case this shift is desirable, it might be convenient to reform other aspects of a country’s fiscal federalism framework. As a result, it is common for property tax reforms to be part of a comprehensive package on fiscal federalism reforms (Box 4.7, on Italy’s 2009 property tax reform, presents an interesting example of a joint reform of property taxation and fiscal federalism).

Revenue equalisation can be employed to offset differences in tax revenue capacity – in this case it is important that upper levels of government prevent local governments from gaming the transfer system

Regarding horizontal relations within levels of government, property tax reforms can lead to an asymmetrical distribution of tax revenues across jurisdictions. This asymmetry can start with the reform’s implementation and may grow over time in line with regional/local characteristics such as the regional/local housing market and administrative capacity. For instance, Färber et al. (2013[36]) found evidence that a reform of the German property valuation system would considerably change the relative tax capacity across Germany’s states. Similarly, Ahmad et al. (2019[37]) simulated the impact of the introduction of a recurrent property tax in China and concluded that the distribution of revenues would be asymmetrical across and within provinces with bigger cities, like Guangzhou and Shanghai, raising significantly more revenues than smaller ones.

Due to these differences in tax revenue raising capacity across jurisdictions, it is common for OECD countries to establish a or to enhance their existing fiscal equalisation system, for a detailed view in equalisation systems see (Dougherty and Forman, 2021[38])). Fiscal equalisation is a transfer of fiscal resources across jurisdictions with the aim of offsetting differences in revenue raising capacity and public service costs. This arrangement is adopted in most OECD countries. According to OECD (2020[13]), these funds, when considering both revenues and costs equalisation together, can amount to roughly 20% of the total subnational expenditure (e.g. in Japan, Austria, Ireland, Korea and Germany). In that regard, when property tax reforms create imbalances on revenue capacity, changes in the fiscal equalisation framework can be adapted to offset these re-distributional effects (see Box 4.7, above and Box 4.8, below).

Exemptions can also cause vertical or horizontal imbalances in property tax revenue capacity. In that case, exemptions granted by lower levels of governments should be treated differently than exemptions granted by upper levels of government. In the first case, the lower levels of government decided to give away their potential tax base and, thus, the potential tax base for equalisation purposes should include the exempted base (i.e. maintaining the revenue capacity of the SNG that granted the exemption identical to its revenue capacity before granting the exemption and, thus, keeping its equalisation fund share constant). In the second case, the exempted tax base should be excluded from the tax base used for computing the revenue capacity for equalisation purposes. That’s because lower levels of government should not be harmed by a decision from upper levels of government, whose impacts are proportional to the asymmetrical distribution of tax-exempt properties across jurisdictions. In this manner, jurisdictions that are losing more revenue capacity are also going to be compensated more by the equalisation system. Lastly, exemptions granted by upper levels of government can also create vertical imbalances, creating resistance from lower levels of government to the exemption. For instance, in Poland local governments asked for compensation for their loss in property tax revenue capacity and demanded that the introduction of new centrally imposed exemptions should require local government consent (Swianiewicz and Lukomska, 2015[39]).

It is important for equalisation transfer systems to be designed in a way that they incentivise tax collection and prevent governments from obtaining a larger share of the equalisation funds than they deserve. In many countries, equalisation penalises tax collection (Blöchliger, 2015[4]) – that is, subnational governments can increase equalisation funds by under-taxing local taxpayers. Most countries minimise this issue by cutting the link between equalisation funds and actual subnational tax revenues through the use of revenue capacity instead of actual revenues in the equalisation formula. Some countries go even one step further by employing a system that rewards subnational governments for collecting more property tax revenue.

Some interesting examples of equalisation systems in which there are incentives for local governments to raise property tax revenues can be found in Norway, Germany, Ireland, among others. Property tax actual revenues of Norway’s municipalities are not considered in the equalisation formula, while in Germany only 64% of municipal property tax revenues are included in the equalisation formula. In Ireland, a local property tax on residential properties (LPT) was introduced in 2013, in addition to the existing property tax levied commercial properties (commercial rates). 80% of LPT’s revenues are retained locally for the funding of basic public services, while the remaining 20% are allocated to the equalisation fund (Slack and Bird, 2014[3]).

In case equalisation funds are based on revenue raising capacity, upper levels of governments should also hinder subnational governments to game the system by systematically under-assessing properties. One way to mitigate this problem is by ensuring that all jurisdictions maintain a specified average “assessed to market value ratios”. Another way is to accept differences in this ratio and adjust the equalisation formula so that property values used for equalisation purpose take into account these differences. These measures minimise distortions in the equalisation base that are caused by differences in valuation policies.

Bundling property tax reforms with reforms in fiscal federalism can be challenging – characteristics of the fiscal system of the country should be carefully analysed

Lastly, despite the fact that bundling a property tax reform with reforms in fiscal federalism can yield benefits, the complexity of the reform can increase substantially. Fiscal federalism reforms are complex and need to be tailored to socioeconomic and institutional characteristics of a country11 (Kim and Dougherty, 2018[40]). There are a myriad of fiscal federalism reforms, ranging from, for instance, a simple change in equalisation rules (e.g. Finnish case) to a more overarching change in subnational expenditure responsibilities and revenue assignments (e.g. Italian case). As a result, designing such a reform is far from trivial. It can also be a strategic choice to separate these two types of reforms (i.e. tax and fiscal federalism) in case this separation would facilitate the approval of important parts of the reform.

In China’s case, special attention should be given to vertical and horizontal disparities

Although China’s regional and urban-rural disparities are declining, they are still significant and higher than the average OECD country. First, OECD (2019[9]) shows that regions differ substantially in many respects, such as medical staff per capita, infrastructure, and disposable income. Second, the Chinese tax system is designed in a manner in which tax yields tend to be more unequally distributed geographically than income. Third, in China vertical disparities between revenues and spending mandates are high, and the gap between own revenues and spending is the greatest at the lowest levels of government (OECD, 2019[9]).

Therefore, a reform that aims to introduce recurrent taxes on immovable property in China can be particularly important to reduce this vertical gap, but it needs to take into consideration these large regional disparities so as to not to further increase them. Such a reform can increase regional inequality for at least two reasons. First, richer local governments are more likely to succeed in the implementation of a well-functioning property tax administration structure that, as was explored throughout this report, has a fundamental importance for collecting property tax revenues. Second, property values are higher in richer jurisdictions, which directly affects their tax revenue capacity. Therefore, in case a fully decentralised property tax system is implemented and local governments do not have the necessary support from upper levels of government, such a tax reform is likely to increase regional disparities in local revenues.

Despite this challenge, OECD countries offer valuable examples of strategies that can be employed to deal with regional inequality. Based on OECD countries’ experience, at least two strategies can be implemented by China: a boost in equalisation transfers and the provision of central support to SNGs. Although China already has in place a substantial equalisation system, the introduction of recurrent taxes on immovable property are likely to increase the disparities in revenue capacity across jurisdictions and, thus, an increase in equalisation funds can be necessary to offset that. It is worth noting that although boosting equalisation transfers is crucial to level differences in local revenue capacity, this solution alone might not be sufficient to level differences in local capacity to implement a sound property tax system. For that purpose, support from the central government can be especially important. This support can be provided in the form of (earmarked) funds, loans or even, in a more extreme situation, a temporary centralisation of the property tax administration. As mentioned before, in case a temporary solution is pursued, having a pre-established date, potentially enforced by law, can help setting expectations and securing that the reform will be implemented fully in the future.