OECD Tax Policy Reviews: Kazakhstan 2020

Report

24 September 2020

4. Equity – sharing the tax burden fairly across the population at large

Abstract

This chapter examines equity issues in Kazakhstan and how the tax burden might be shared more fairly across the population. It focuses on personal income taxes, social security contributions and value added taxes. Recommendations to strengthen the design of these taxes are provided in the table overleaf.

Box 4.1. Recommendations to strengthen to design of the PIT, SSCs and VAT

PIT

PIT design

Strengthen the capacity of tax administration to monitor incomes and income sources to support the introduction of an end-of-year tax declaration.
Introduce the end-of-year tax declaration as soon as administrative feasible to support the introduction of a progressive PIT system; do not continue to defer its introduction.
To raise the currently low levels of PIT revenues, introduce a progressive PIT rate schedule over the medium-term.
Turn the basic tax allowance into a basic tax credit and design the progressive PIT rate schedule such that the 90% exemption of income for a targeted group of taxpayers becomes unnecessary.
As a short-term measure, consider increasing the flat PIT rate modestly.
For any PIT rate reform, take account of the overall tax wedge, which is rising due to the proposed increased SSC rates.
Before introducing a progressive PIT system, develop a PIT microsimulation model using individual-level tax return to more robustly cost PIT reform proposals and measure distributional impacts.

Personal capital income tax design

Abolish the exemptions on personal capital income including the exemption for capital gains from the sale of shares held for more than 3 years and for government securities, bonds and shares officially listed on a Kazakhstan stock exchange.
Replace the current set of differentiated rates on different forms of personal capital income with a proportional capital income tax rate or a progressive tax rate schedule. Consider taxing personal capital income jointly with labour income, or introduce a separate progressive tax rate schedule for capital income at the individual level

SSCs

Continue with the reform of the SSC system

To support the financing of the welfare system with increased SSC funds, continue to implement the reform for a new SSC system.
Maintain the employer and employee SSC rates on insurance, health and pension at the currently proposed rates but do not increase the SSC rates further so as not to add further to the rising employment costs for firms.

The SSC base is narrow by design and could be broadened

Consider reviewing the atypical deduction of pension contributions from the SSC base.
To raise SSC funds, consider abolishing the atypical and expensive inheritability of pension payments.

Other SSC reform considerations

Consider partly financing social benefits through taxation (in the medium or longer run when the role of the PIT in the tax system is strengthened).
Implement less frequent but more consistent SSC reform.

VAT

Consider increasing VAT rates

Consider increasing the standard VAT rate, which is low by international standards, but any VAT rate increase should be modest given the inflationary environment.
Continue to strengthen the design and operation of the VAT following international best practice and do not introduce a sales tax.

Broaden the VAT base

Broaden the VAT base by removing various exemptions for goods and services including for example the exemption for newly constructed residential buildings that are brought on the market for the first time.
Fully restore the VAT chain by applying the standard VAT on all transactions to and within the Special Economic Zones (SEZs) (i.e. by including zero-rated supplies for transactions within the standard VAT system).
Build on progress to date by applying simpler and faster VAT refunds.
Continue to lower the VAT registration threshold (conditional upon the capacity of the tax administration to bring more VAT payers into the system).
Consider implementing additional VAT simplification measures such as a VAT flat-rate scheme.
Review the VAT exemption for goods imported by individuals ‘not for entrepreneurial purposes’, which may incentivise avoidance in its current form.

Strengthen VAT compliance

Continue to adapt the VAT rules to increasing digitalisation and online sales. This reform will broaden the VAT base by ensuring the taxation of inbound digital supplies, in line with the OECD International VAT/GST Guidelines.
Consider making digital platforms liable for the VAT on sales by online traders.
Consider adopting new tools to tackle sales suppression and invoice falsification.

Excise, property and carbon taxes

Consider partly financing healthcare with excise duties on alcohol and tobacco.
Raise additional tax revenues from recurrent taxes on residential immovable property in the medium-run; develop a fiscal cadastre and record information related to the market values of property and make the information available for tax purposes.
Envisage the introduction of a carbon tax and increase environmentally-related tax rates.

4.1. To finance the NOD, PIT revenues could be raised through a broad and progressive PIT

There may be scope to finance the NOD by raising the currently low levels of PIT revenues. The PIT-to-GDP ratio in Kazakhstan is 1.4%, which is low compared to the OECD average of 8.2%. As discussed, the non-oil budget deficit (NOD) could remain at as high as 6% by 2025 (under moderate revenue and expenditure conditions). In order to finance this deficit, Kazakhstan could consider modestly increasing the PIT-to-GDP ratio in the direction of the OECD average. This could for example be done by increasing the PIT-to-GDP ratio by between 1 to 2 percentage points to reach approximately 3% of GDP between now and 2025. Such a PIT-to-GDP level would still be significantly less than half of the OECD average.

4.1.1. Personal income tax revenues are low

PIT as a share of total tax revenues are low. Taxes on the income of individuals are low in Kazakhstan, making it an outlier relative to international and regional comparison countries. Low PIT revenues reflect low incomes in Kazakhstan (discussed in 2.1.2). PIT accounted for 8.6% of tax revenues in Kazakhstan in 2017, below the OECD average of 23.8% and the CIS average of 17%. In Kyrgyz Republic, Azerbaijan and Russia for example the PIT share is higher than in Kazakhstan.

Figure 4.1. Taxes on individuals are low by international and regional standards

PIT as a % of total tax revenues, selected countries including from OECD and CIS, 2017

PIT revenues have been declining slowly in recent decades. Over the period 1998 to 2016, PIT revenues as a share of GDP have declined slowly. More recently, and since 2012, PIT revenues have plateaued at a rate of just below 1.5% of GDP. While a greater degree of volatility is seen in PIT as a share of total revenues, this arises largely due to the associated volatility in total tax revenues. During the period 2002 to 2006, when a four-bracket progressive PIT was in place with rates ranging from 5% to 30%, PIT as a share of tax revenues declined (Institute for Economic Research in Kazakhstan). The current flat PIT rate system was introduced in 2007. However, in the years immediately prior to and after 2007, PIT as a share of tax revenues remained relatively flat indicating that the new system might have had little discernible impact on tax revenues (discussed further below).

Figure 4.2. PIT revenues have been declining slowly over time

PIT revenues as a proportion of tax revenues and GDP in Kazakhstan, 1998 - 2016

4.1.2. The effective tax burden on labour income remains relatively low

The implicit tax rate on labour is similar to the lowest rate countries in the EU. The effective tax burden on labour can be approximate in various ways. One method is the implicit tax rate (ITR) on labour, which is an overall aggregate indicator used in the national accounts. The ITR is a so-called back-ward looking indicator as it uses information from the past. The ITR on labour can be defined as taxes and SSCs on employed labour income divided by total compensation of employees and payroll taxes. Under the OECD revenue statistics definition, payroll taxes are largely comprised of the social tax in Kazakhstan1 In Kazakhstan the ITR on labour is 24.5% in 2018.2 The ITR on labour in Kazakhstan is similar to some of the lower rate countries in the EU such as Bulgaria and Malta (24.3% in both) and the United Kingdom (25.9%) in 2017 but lower than the EU-28 average which was 36.3%.

Forward-looking tax wedge indicators confirm that the average tax burden on labour income in Kazakhstan remains significantly below the average in the OECD. A further way to measure the effective burden on labour is to estimate the average tax wedge, which is a forward-looking indicator that models the combined effect of tax rates and rules (including SSCs) for workers who earn a particular income. The tax wedge can be defined as PIT, total SSCs and payroll tax as a proportion of labour costs (where labour costs is gross income plus employer SSCs plus payroll taxes). The tax wedge differs from the ITR in that it includes employer SSCs in the denominator and does not use National Accounts information but makes assumptions on the income that is earned by the worker.

Figure 4.3. The average tax wedge on labour income in Kazakhstan remains relatively low compared to the OECD average

Tax wedge, single person at 100% of average earnings, 2018

4.2. New PIT reforms are being introduced that will narrow the PIT base

4.2.1. Key design features of the PIT

Personal income tax is levied on a worldwide basis in Kazakhstan, which means that residents are taxed on their worldwide income. Non-residents are subject to tax only on their income sourced in Kazakhstan. This is the most common system internationally (Shum, Fay and Man Ching, 2017[1]).

The flat-rate PIT system has not altered PIT revenues in the years immediately following its introduction. Having previously had a progressive PIT rate system, Kazakhstan shifted to a 10 percent flat PIT rate in 2007. A main rationale for the reform was to simplify the PIT system. Based on a simple visual inspection of PIT revenues over time, the flat-rate reform does not appeared to have significantly altered PIT revenues in the immediate years following its introduction. However, there are challenges to drawing causal interpretations from a simple before and after inspection of PIT revenues. To illustrate this with an example, in 2001 Russia replaced its higher PIT rates with a flat PIT rate of 13% and PIT revenues increased by 26% in the following year. However, an econometric analysis of the reform demonstrated that the PIT revenue increase was not attributable to the flat-tax reform but rather to other changes occurring at the same time such as measures that strengthened the capacity of the tax administration (Ivanova et al., 2005[2]).

Similar to most OECD countries, a minimum amount of income is exempt from PIT through a personal basic allowance. The main personal allowance for employment in Kazakhstan is the basic allowance, which allows employees to deduct one Minimum Monthly Salary (MMS) from gross income each month. The total amount of tax deduction per year must not exceed 12 MMS (USD 1 315) over a calendar year. While this structure suits employees who have a consistent regular monthly income, it may penalise employees who have no income in some months of the year.

The basic allowance is high having been increased recently. The amount of the allowance has been increased in recent years along with increases in the MMS. The MMS increased moderately between 2016 (KZT 22 859) and 2018 (KZT 28 284) before increasing sharply in 2019 (KZT 42 500). The rate of increase has outstripped inflation. For example, on average between 2016 and 2019, the basic allowance increased by 20% annually while inflation was 8.4% annually over the same period.3 The current basic allowance is relatively high, and represents about one-quarter (26%) of the mean wage, half (52%) of the median (KZT 82 977) or one-fifth (18%) of the top 10% of employment incomes.4

4.2.2. Proposed PIT reforms

Plans exist to introduce an end-of-the year tax declaration for individuals, which would align the administration of the PIT with common practice in the OECD. In recent years, the authorities have announced plans to introduce an end-of-the-year tax declaration whereby the tax administration would add all sources of income earned throughout the year to determine the individual taxpayer’s end-of-the-year tax liability. The taxes that are withheld during the year would be deducted from the final tax liability and individual taxpayers would be required to pay the remaining tax liability to the tax administration or they would receive a tax refund in case too high taxes have been withheld during the year.

However, the proposed plans have been repeatedly delayed. Most recently, the end-of-the-year tax declaration was due to be introduced in 2020. However, its introduction has since been delayed until 2025. While an end-of-the-year tax declaration increases the burden on the tax administration and, therefore, requires strengthening the capacity of the tax administration, the reform would bring many advantages. Government should therefore implement the necessary measures such that the end-of-the-year tax declaration can be introduced as soon as possible. This will involve strengthening the capacity of the tax administration and in particular its IT system as well as hiring and training the staff of the tax administration to successfully perform the new tasks.

An end-of-the-year tax declaration is a necessary condition for the PIT to play its key role in the tax system. The introduction of the general tax declaration is a necessary foundation for any successful introduction of a progressive PIT rate schedule, as it will allow applying the progressive rate schedule on the combined sources of income earned throughout the year. In addition, the reform would allow government to achieve broader objectives by using the tax system (e.g. through the introduction of measures to alleviate the burden on families with children). The general tax declaration will complement third-party reporting and the current withholding tax system, which should be maintained.

4.3. Towards a progressive PIT system to enhance fairness

4.3.1. General PIT reform considerations

Broadening the tax base is preferable to increasing tax rates. Broad tax bases and low tax rates is considered best tax policy; the arguments have been well documented in other reports and will not be repeated here. In countries with somewhat high informality such as Kazakhstan, raising revenues through tax rate increases comes with additional risks. In Kazakhstan, 90% of PIT comes from employees with only 9% from self-employed, despite widespread self-employment, indicating possibly high informality (according to research from the Institute of Economic Research in Kazakhstan). Raising tax rates on the formal sector would reinforce the distortions between the formal and the informal economy; it would strengthen the incentives for businesses and workers to operate in the informal economy. In contrast, broadening the tax net to a larger number of taxpayers makes it possible to collect more revenues while avoiding that tax rates would have to be increased on the formal sector. Informal workers, defined as those who do not pay SSCs and unregistered self-employed, account for 20% of total employment in the country by some estimates (OECD, 2017[3]). A discussion on the general consequences of informality are provided in Box 4.2.

Kazakhstan could turn its basic tax allowance into a tax credit. Many OECD countries have replaced their tax allowances with tax credits. The main argument in favour of tax credits, which are deductible from tax liability (unlike tax allowances, which are deductible from taxable income), is that the value of tax credits is independent of income while the value of tax allowances is increasing in the taxpayer’s marginal tax rate if PIT rates are increasing with higher incomes. Therefore, unlike tax credits, tax allowances typically benefit those on higher incomes more. If Kazakhstan introduces a progressive PIT system, discussed later in this Chapter, it could simultaneously turn its basic allowance into a tax credit to support fairness across the income distribution.

Box 4.2. The consequences of informality

Informality generates greater inequality. Workers employed in the informal sector have limited access to social protection, relatively lower wages and are vulnerable when they lose their job or retire. High levels of informality may also reduce workers’ access to training, which can worsen skills shortages, leading to greater societal inequalities.

Informality affects productivity and growth. Informal sector production can generate inefficiencies, either because companies limit their size below their optimal efficiency scale to avoid being detected or because they use outdated production technologies (Andrews, Caldera Sánchez and Johansson, 2011[4]). The relative cost advantages enjoyed by informal companies may allow them to stay in business even if they are not productive (Andrews, Caldera Sánchez and Johansson, 2011[4]). However, companies operating in the informal sector also have a more limited access to finance which constrains investment and to qualified labour.

Informality reduces tax revenues. In addition to informal workers not paying taxes, many may also receive social benefits, adding to the state fiscal burden. In Kazakhstan, the informal sector tends to have low incomes however which means that the tax revenue loss is likely to be limited.

Informality erodes trust. Finally, high levels of informality, when observed by formal workers, can result in an erosion of trust in public institutions and result in lower tax morale, which may lower revenues through other channels. Importantly, the larger the informal sector, the more incentives people have to remain or become informal, particularly if there is a view among some that the informal sector is tolerated.

Source: OECD analysis.

4.3.2. After the new PIT reform of 2020, a significant numbers of low-income earners will pay little to no PIT

A new reform will exclude 90% of taxable income for low-income taxpayers. To reduce the tax burden on low-income employees, Kazakhstan introduced a new reform in 2020 that will exclude 90% of taxable income from PIT for employees with gross income up to 25 MCI (or KZT 63 125). Those on higher gross incomes above the threshold do not benefit from the income exemption. This tax exemption comes on top of the relatively high basic tax allowance as well as the deduction of pension contributions from taxable income; both measures together imply that taxpayers with monthly gross income up to KZT 48 000 do not pay PIT.

The reform reduces average effective PIT rates significantly, but albeit for a narrow-range of incomes only. The reform implies that an employee with income of KZT 63 125 will pay an effective average PIT rate (on gross income) of one-fifth of one percent (0.2%). The tax due and effective tax rates are calculated by first taking into account the deduction for the pension contributions and the basic tax allowance; the remaining taxable income is then reduced by 90%. For example, gross income equal to KZT 63 125 will be first reduced with the 10% obligatory pension contribution (KZT 6 312), and then with the basic tax allowance (KZT 42 500). The remaining taxable income of KZT 14 312 will be reduced by 90% to KZT 1 431 and then the 10 percent PIT rate is applied. This results in an average effective tax rate of about 0.2%. Figure 5.4 shows that the reform will only affect the effective PIT rate for a small band of employees with incomes between KZT 48,000 and KZT 63 125, as workers with monthly gross earnings below KZT 48 000 do not pay PIT as a result of the deduction of pension contributions and the basic tax allowance. As a result, the reform is likely to affect roughly 11.2% of employees, based on the limited picture that grouped distributional data can provide.

The reform might be distortive. Unlike in the case of increasing the basic allowance, the reform to exclude 90% of taxable income for low earners exempts a large part of income from PIT for low-income workers without giving any tax reduction to those on higher incomes. Therefore, relative to other reforms options such as the basic allowance,5 the reform is more equitable because it provides no advantage to those on higher incomes. However, it is relatively less efficient because of its potential to distort behaviour because it produces a high marginal effective tax rate (METR) at the ceiling of KZT 63 125, which reduces the incentive for taxpayers to earn more income beyond that point.

The reform may incentivise employees to keep income below the maximum ceiling and to have multiple jobs to reduce PIT liability. The reform to exclude 90% of taxable income for low earners will produce a change in the effective and marginal effective PIT rates above and below the 25 MCI ceiling (Figure 4.4). For example, an employee earning KZT 63 000 will pay KZT 142 while an employee earning an additional KZT 1 000 (KZT 64,000) will pay more than ten times the PIT (KZT 1 510). From an incentives perspective, this encourages employees to report income just below the threshold, known as ‘bunching’. This could occur for example if the taxpayer did not declare income past this amount (for instance by working part-time in the informal economy). It could also occur through a reduced incentive for employees to work more hours or find better-paid employment. Many countries have found evidence of individual taxpayers and companies ‘bunching’ below points of discontinuity (kinks or notches) in tax systems (Boonzaaier et al., 2017[5]) (Kleven, 2016[6]). However, despite the large jump in the tax rate at the threshold, the overall actual amount of PIT paid remains small relative to income, which may limit the extent of this effect. Furthermore, if the reform applies separately across jobs, it may incentivise employees to take two jobs at incomes below the threshold to avail of a lower PIT liability.

Figure 4.4. The reform may encourage some employees to report income below the ceiling

Average effective PIT rate in 2018 and 2019 by income threshold

The standard approach to increasing PIT progressivity would be to replace the flat PIT rate with a progressive PIT rate schedule where higher tax rates in higher taxable income brackets. Most OECD countries apply progressive PIT rate schedules. Indeed, the PIT is the main source of progressivity in the tax system in most OECD countries. The advantage of a progressive PIT schedule reform is that it could increase progressivity and PIT revenues simultaneously by increasing PIT on the highest earning employees while at the same time preventing peaks in marginal effective tax rates, as is the case with the 90% taxable income exemption in Kazakhstan. The introduction of a PIT rate schedule would need to go hand in hand with the end-of-the year tax declaration, as pointed out above. In the short run, Kazakhstan could consider replacing the 90% exemption of taxable income for low-income workers with an additional tax allowance that would be decreasing for higher incomes. If well designed, such an additional tax allowance would reduce tax liabilities for taxpayers that benefit from the 90% exemption but, as it tapers out for higher incomes, would avoid the sharp peak in effective tax burden at the eligibility threshold.

4.3.3. Simulating the introduction of a progressive PIT rate schedule

This section produces various simulations to estimate the potential impacts of PIT rate and base changes in Kazakhstan. The analysis will evaluate the introduction of alternative PIT reforms. The analysis is for illustrative purposes. It could have been carried out for alternative designs of the PIT rate schedule.

The analysis is based on the grouped monthly income distribution data for the year 2017, shown previously in Figure 2.5. To recap the summary statistics from these data, the mean income is KZT 16 673 (USD 423), the top 10% is KZT 236 350 (USD 615) and the bottom 10% is KZT 27 235 (USD 71). For the purposes of the analysis, the number of employees is based on data from a study by the (Institute of Economic Research, 2018[7]), which shows that there are 8.6 million employees in 2017.

The simulation analysis is based on a number of methodological steps. The methodological approach taken involves the estimation of monthly average incomes, the calculation of effective PIT rates and finally the estimation and evaluation of total PIT paid. These are described in turn below.

1. Estimation of monthly average gross and taxable incomes. First, a midpoint monthly income is calculated for each income category (for example, KZT 82 500 is approximately the midpoint of KZT 75 000 and KZT 90 000). Total monthly income in each category is estimated by multiplying the midpoint income by the number of employees. For example, 9% of employees have monthly income between KZT 20 000 – 30 000, so assuming all employees earn the midpoint income in this range of KZT 25 000 and then multiplying by 12 months gives annual income of KZT 237 million for all of the employees in that particular income category. Total annual income is estimated by following the same approach across all categories and summing them together. Next, average monthly taxable income is calculated at each category by subtracting the obligatory pension payment (10% of gross monthly income) and the basic allowance from the monthly gross income.6

2. Calculation of effective PIT rates. The effective PIT rate is calculated at each income category by applying the progressive PIT rates and brackets to the midpoint monthly taxable income. For example, applying the progressive rate schedule and brackets from Table 4.1, the employee will pay PIT of 1% on the first KZT 42 500, 10% on the next KZT 57 500 up to KZT 100 000 and so on. The same approach is applied to all income categories to give new estimates of PIT and the effective PIT rate at the employee level.

3. Estimation and evaluation of total PIT paid. Finally, to estimate total PIT paid in Kazakhstan in each income category, the calculated effective PIT rate for the hypothetical employee is applied to the estimate of total annual income. This modelling approach can be evaluated by comparing estimated and actual PIT revenues. Applying the flat PIT rate to total taxable income gives a total PIT estimate of KZT 784 817 million, which is less than actual PIT of KZT 838 394 million in 2017 (based on data from the State Revenue Committee).

A number of modelling scenarios are considered. The simulation analysis is used to model the potential impact of various reforms on effective tax rates and tax revenues. The modelling scenarios considered include the following:

The level of the basic allowance and the impact of the 90% taxable income exemption;
Increases in the flat PIT rate from 10% to 15% and 20%; and
The introduction of a progressive PIT rate schedule shown in Table 4.1.

Bracket (KZT)	Progressive PIT schedule
Bracket 1 (up to 45 000)	5%
Bracket 2 (455 00 – 100 000)	10%
Bracket 3 (100 000 – 200 000)	15%
Bracket 4 (200 000 – 300 000)	20%
Bracket 5 (300,000 and above)	25%

Source: OECD analysis.

The simulation analysis comes with caveats due to limited data. The simulation analysis presented has several caveats. First, from an analysis perspective, the precise income of individuals was not available to the authors and it is assumed that taxpayers in each category earn the midpoint monthly income. This assumption means that the PIT simulation are based on an approximate rather than actual income distribution. For example, the broad nature of the grouped distribution data mean that small changes in the effective rate, for example related to bunching activity, should be interpreted with considerable caution. The analysis also assumes no behavioural change arising as a result of the tax rate changes. In addition, the analysis is restricted to employees (self-employed are not included) and consideration is not given to SSCs and allowances other than the basic allowance and compulsory pension contribution. Notwithstanding these methodological limitations, the averages calculated based on the grouped data are similar to the actual averages in the distribution. For example, the mean monthly income based on the grouped mid-point income data is KZT 174 000, which is similar to the actual mean of KZT 162 673. Overall, the approach provides suggestive but not conclusive evidence of the likely direction and magnitude of PIT revenues arising from moving towards a progressive income tax schedule.

Kazakhstan could consider developing a microsimulation model to cost reform proposals. Kazakhstan could consider developing a microsimulation for individuals by combining individual-level tax return, SSC and survey data. Such a simulation model would have the potential to rigorously cost different PIT and SSC reform options along with the likely distributional impacts.

The increased basic allowance has reduced effective average tax rates for all taxpayers

The increased basic tax allowance reduces the effective tax rate on all employees. In general, basic tax allowances are regressive in the sense that all taxpayers benefit from them. In particular, taxpayers that face higher marginal statutory tax rates benefit more in nominal terms. Figure 4.5 shows the simulated average effective PIT rate across the income distribution based on increases in the basic allowance in 2017 (when it was KZT 24 459), 2018 (KZT 28 284), 2019 (KZT 42 500) and finally in 2019 including the both the basic allowance and the 90% taxable income exemption. For each of these four scenarios, the results show that the overall average effective PIT rate across the distribution falls from 7.60% to 7.40% to 6.75% to 6.73% respectively. The increased basic allowances in 2018 and 2019, but particularly in 2019, had the largest impact on reducing the overall effective PIT rate. The increased basic allowances also have the general effect of shifting the effective PIT rate downwards across the distribution but not changing the shape of the distribution by too much. It can also be seen that the average effective PIT rate increases with higher incomes – in 2019 for example, the average effective PIT rate ranges from 1% at low incomes up to 7.9% at higher incomes7. Overall, the increased basic allowance will reduce PIT paid and increase the take-home pay for all employees while also increasing the point at which employees begin to pay PIT.

Increases in the basic allowance have eroded PIT revenues. Figure 4.6 simulates the PIT revenue reduction associated with increasing the basic allowance in 2018 and 2019 relative to the 2017 basic allowance scenario (i.e. taking into account the flat 10% PIT rate). The results are shown for each of the selected income brackets shown in Table 4.1. Overall, relative to the 2017 basic allowance scenario, the PIT revenue loss associated with the 2018 allowance, the 2019 allowance and finally the 2019 allowance plus the 90% taxable income exemption is estimated to be about 4%, 19% and 20% of total 2017 PIT revenues, respectively.

Figure 4.5. The increased basic allowance reduces the effective tax rates on all taxpayers, not just those on low incomes

Estimated average effective tax rate changes arising from increased basic allowance between 2017 and 2019

The increased basic allowance narrows the PIT base by removing millions of low-income taxpayers from the PIT net. The effects of introducing the basic allowance are likely to be substantially different across the income distribution and the selected income brackets shown in Table 4.1. Based on the 2017 allowance, some low-income taxpayers in the first bracket (up to KZT 45 000) would have previously paid some PIT. The simulation shows that the combined effects of the increase in the basic allowance in 2018 and 2019, plus the taxable income exemption reform, fully reduce PIT revenues from those earning less than the lowest bracket of KZT 45 000 and further reduce PIT revenues by over 40% from those in the second bracket. In the third bracket, where some 40% of all PIT is paid based on the distribution data, the simulation analysis indicates that there could be a further 16% revenue loss.

The increased basic allowance will blunt the impact of the new taxable income exemption reform. Of the recent basic allowance increases, the 2019 allowance produces the largest PIT loss (Figure 4.6). The analysis also show that once the 2019 basic allowance is in place, the additional impact of the 90% taxable income exemption reform is minimal because many of the same low-income taxpayers are already benefiting from the substantial increased in the basic allowance, as pointed out before.

Figure 4.6. Recent reforms will significantly reduce PIT revenues

Estimated cumulative loss in PIT revenues (base scenario is basic allowance for 2017) taking into account the income distribution in 2017

Introducing a new progressive PIT system could increase PIT revenues while supporting equity

Increasing the flat PIT rate would increase the tax burden across the income distribution. Figure 4.7 shows simulations for different flat-tax rates of 10%, 15% and 20% under the 2018 basic allowance, the 2019 basic allowance and the 2019 basic allowance under the taxable income exemption. First, and as expected, for all levels of the basic allowance, increasing the flat PIT rate to a higher 15% or 20% will have the effect of increasing the tax burden across the income distribution thereby increasing PIT revenues. Second, the analysis highlights how a combined flat rate and basic allowance produce progressivity in the PIT system, as is already the case in Kazakhstan. Third, and as demonstrated in Figure 4.5, the effect of the increased allowance in 2019 and the 90% taxable income exemption in excluding increasingly larger shares of taxpayers from the PIT net can also be seen by looking across the three graphs in Figure 4.7.

Figure 4.7. Flat PIT rate simulations

Simulations of EATR under different flat PIT rates and basic allowances and exemptions

Introducing a progressive PIT rate schedule could increase progressivity by shifting the tax burden from lower to higher income employees. Under the current system in Kazakhstan, the flat PIT rate, basic allowance and the 90% taxable income exemption effectively produce progressivity in the PIT system while exempting low-income taxpayers from paying any PIT. However, an alternative approach is also possible. Figure 4.8 shows a proposed progressive income tax schedule based on a rate schedule of 5%, 10%, 15%, 20% and 25% (shown in Table 4.1), while maintaining the basic tax allowance at its 2019 level of KZT 42 500. The analysis shows that, compared to the current flat-rate system, a progressive PIT rate schedule has the potential to further shift the tax burden from lower to higher earners while maintaining PIT revenues at about the same level (assuming no behavioural change).

Figure 5.9 shows that the increase in PIT revenues from the second-highest and highest income brackets, those with incomes between KZT 200 000 and KZT 300 000 and in excess of KZT 300 000 respectively, could be as much as 12% and 42% respectively relative to the current flat rate with the basic allowance. Taxpayers that fall in the third bracket would on average pay less PIT because the low 5% PIT rate (instead of the 10% flat rate) in the first tranche of taxable income reduces their tax liabilities more than the higher 15% PIT rate in bracket 3.

Figure 4.8. Introducing a progressive PIT rate schedule could increase progressivity while maintaining current levels of PIT revenues

Estimated average effective tax rate changes arising from a progressive PIT rate schedule

Figure 4.9. A progressive PIT schedule would shift the tax burden from lower to higher earners

Estimated change in PIT revenues from a progressive PIT rate schedule (base scenario is current flat-rate with basic allowance 2019)

Before implementing a progressive PIT system, limitations and risks should be addressed including the introduction of the universal tax declaration to allow full visibility on the incomes of all citizens. There are however several challenges with regard to moving to a progressive PIT system in the short-term given the current income distribution in Kazakhstan. For example, a study by the Institute of Economic Research shows there is a significant 33% of workers classified as ‘working poor’ in Kazakhstan in 2017 and there is a large gap between the minimum and median wage. In addition, research on the possible transition to a progressive PIT system highlights potential risks including a reduction in local budget revenues with low wages, increased costs for tax administration and tax evasion (Alpysbaeva et al., 2020[8]). As mentioned earlier, greater clarity on the extent of these challenges could be gained through the development of a PIT microsimulation model using individual-level tax return data. Therefore, the transition to a progressive PIT system should be considered over the medium-term and not before these risks and limitations are understood and addressed.

Kazakhstan may also want to re-evaluate the rules that define when an individual is resident for income tax purposes. According to article 217 of the Tax Code, tax residents are individuals who permanently reside in Kazakhstan or who non-permanently reside in the Republic of Kazakhstan but whose centre of vital interests is located in Kazakhstan. The rules that define the latter conditions, in particular regarding the ownership of immovable property, might have to be revised and aligned with international best practice. An in-depth analysis is left for further work

4.3.4. Personal investment income is mostly exempt from tax in Kazakhstan

Many common forms of personal capital income are exempt from tax in Kazakhstan and where personal capital income is taxed, tax rates are low. Personal capital income is largely exempt from tax in Kazakhstan because of a generous set of exemptions for income generated in the form of dividends, capital gains and interest. Capital gains from the sale of shares of Kazakh companies are exempt provided the shares are held for 3 years or more and the company fulfils certain criteria. In addition, government securities, government bonds, and shares officially listed on a Kazakhstan stock exchange are exempt from capital gains tax. Interest from shares issued by government are exempt from taxes on income, as are interest on bank deposits and debt securities. Where personal capital income is taxed, the rates are low. For example, the dividend tax rate for residents is only 5% and the capital gains tax rate is 10%.

Low effective tax on personal capital income may distort behaviour and create avoidance opportunities. As capital income is taxed at low and often zero effective rates, it is taxed more lightly than labour income and the difference is larger for high-income employees whose labour income is taxed at relatively higher effective tax rates. This could create economic distortions and raises questions of equity, particularly for low-income taxpayers who may be less likely to generate capital income. This may provide an incentive for an employee to setup a company in a non-subsurface area and pay themselves through untaxed dividends.

Compared to OECD countries, marginal effective tax rates (METRs) on savings in Kazakhstan are low. The OECD calculates METRs on household savings to assess the impact of a wide range of taxes and tax design features on the incentives to save in different assets (OECD, 2018[9]). The METR calculations take into account a range of taxes levied on household savings, deductions and variations in the tax base, different asset holding periods and the potential build-up of untaxed or tax-deferred returns. METRs also incorporate the impact of inflation, which can impose a substantial additional tax on the return to savings. Figure 5.10 presents METRs on bank deposits, government bonds and shares in Kazakhstan compared to the unweighted OECD average. Due to differences in average rates of inflation between Kazakhstan and OECD countries, inflation is set to zero to improve comparability. The exemption of common savings vehicles from personal income taxation discussed above results in an effective tax rate of zero across most savings vehicles. Even when the holding period for shares does not reach the 3 years required for a tax exemption, the marginal effect tax rates in Kazakhstan are low compared to the OECD average.

Figure 4.10. Marginal effective tax rates on savings in Kazakhstan are low

Marginal effective tax rates (METR), 100% of the average wage, inflation 0%; interest rate 3%; holding period 6 months

Taxes on personal capital income could be increased by expanding the tax base and introducing a single flat rate on all forms of capital or a progressive rate alongside a progressive PIT rate schedule that applies to labour income. The taxation of personal capital income varies substantially in OECD countries. Some countries tax all personal capital income at a flat rate and wage and pension income at progressive rates (dual income tax systems) and other countries tax all or most capital income together with labour income at a progressive PIT rate schedule (comprehensive income tax systems) (OECD, 2004[10]). Given that most forms of capital income are exempt from tax, and where it is taxed rates are low compared to OECD counties, Kazakhstan could consider expanding the capital income tax base and increasing tax rates. This could for example include removing some of these exemptions and introducing a single flat tax rate on all forms of capital income rather than the current differentiated set of rates (for example, where dividend tax rate for residents is 5% and capital gains tax rate is 10%). Alternatively, and if the year-end tax declaration and the progressive PIT rate schedule are introduced, Kazakhstan could choose to tax capital income together with labour income. A further option could be to tax capital income at a separate progressive capital tax rate schedule, where the top rates should be sufficiently high to reflect that capital income is generally earned by higher-income individuals.

Automatic exchange of information (AEOI) provides opportunities for Kazakhstan to strengthen the fairness of its tax system. Previously, tax administrations had to request information from their residents on capital income earned in other jurisdictions (exchange of information on request). Under the new automatic exchange of information rules, it has become more difficult for an individual to conceal capital income abroad and tax administrations have become more effective at verifying compliance. In the future, this could present opportunities for Kazakhstan to increase the level of taxation of capital at the individual level. However, the level of taxation of capital at the individual level is low and almost all foreign assets held by Kazakh residents is in real estate. KZT 62.2 billion is held in real estate by Kazakh residents compared to KZT 5.1 billion in shares and KZT 2.9 billion in securities according to data from the Ministry of National Economy.

4.4. Financing the welfare system through Social Security Contributions

While the current reform of the SSC system in Kazakhstan will help to raise financing for the underperforming health and welfare systems, options exist to broaden the base (OECD, 2017[11]).8 The SSC-to-GDP ratio is 0.5% in Kazakhstan, which is very low compared to the OECD average of 9.2%. Furthermore, SSCs account for 3.2% of total tax revenues compared to 26.2% for OECD countries. However, these figures do not take into account the employer payroll tax (i.e. the social tax). Currently in 2020, Kazakhstan is in the process of reforming its SSC system which includes the introduction of new SSCs while also increasing current SSC rates. This reform will go some way to shifting the SSC-to-GDP ratio to align more closely with the OECD average, which will provide much needed financial resources to support the underperforming health and welfare systems. The timing of the reform may be prudent given Kazakhstan’s significant demographic advantage compared to many other developed and OECD countries. It has a large and expanding working-age population, which is an opportunity for the country, and the old-age dependency rate remains low and is not yet a significant challenge relative to many other countries. However, the SSC base remains narrow by design. To take two examples which are atypical internationally, pension contributions are deductible from the SSC base and pension payments can be inherited. Such policies are generous but expensive for the State. A number of alternative policy designs could be considered to increase financial resources by broadening the SSC base. Furthermore, there have been frequent and significant SSC proposed policy changes in recent years, some of which have subsequently been delayed, which produces unnecessary administrative complexity and planning challenges for individuals and businesses. Instead, Kazakhstan could benefit from less frequent but more consistent SSC policymaking.

4.4.1. Increasing the performance of the welfare system will require more financial resources

Despite becoming a policy priority in recent years, health outcomes are poor and health expenditures remain low. Health outcomes in Kazakhstan fall well short of OECD countries, with the average life expectancy more than 6 years below the OECD average. In addition, many countries with similar levels of income such as Hungary, Poland or Turkey, continue to outperform Kazakhstan on health outcomes (OECD, 2018[12]). The average life expectancy at birth is 63.4 years in 2016, compared to an average of 69.7 for the OECD and CIS countries shown in Figure 4.11. The levels of expenditure on healthcare are low in Kazakhstan (Figure 4.12). Health expenditure as a share of GDP are 3.5% in Kazakhstan in 2016 compared to CIS countries (5.3% in Russia, 6.3% in Uzbekistan and 6.7% in the Ukraine) and 12.6% in the OECD. Moreover, public health spending is only 1.8% of GDP, thus contributing only 58% of total health expenditure, and leaving high out-of-pocket costs for patients (OECD, 2018[12]).

Healthcare, social welfare and pension system demands will rise gradually over the longer-term with demographic changes. Populations are ageing rapidly across advanced economies and many emerging market economies because of rising life expectancy and declining fertility (Colin and Brys, 2019[13]). Unlike many OECD and developed countries, Kazakhstan does not have an aging population. However, it should prepare for aging over the next decade as the number of older-people and pensioners begins to rise. Although the country has a low current old-age dependency ratio, it is expected to increase over the next decade, which will come with a rising demand for healthcare. In addition, social welfare demands are rising. Over the past 5 years, the number of recipients of targeted social assistance has grown from 77 000 to 1.4 million. The amount of funds allocated from the budget for social support since 2017 has increased more than 17 times according to the (State of the Nation, 2019[14]).

Figure 4.11. Health expenditure is low in Kazakhstan compared to CIS and OECD countries

Health expenditure as a share of GDP, 2016

Figure 4.12. Despite economic growth, health outcomes remain relatively poor

Life expectancy at birth in OECD and CIS countries, 2016

Compulsory pension contributions have been rising in nominal terms. Pension contributions have been rising in nominal terms in recent years, from KZT 597 billion (USD 1.5 billion) in 2014 to KZT 849 billion (USD 2.2 billion) in 2018, according to data provided by the authorities. However, as a proportion of GDP, pension contributions have remained more stable (1.5% of GDP in 2014 and 1.4% in 2018). 95% of all pension contributions are compulsory pension payments; the remainder is comprised of compulsory professional pension contributions (4.6%) and voluntary pension contributions (0.04%).

Kazakhstan has a key demographic advantage over many countries with its large and expanding working-age population, which is an opportunity for the country. Kazakhstan has a high proportion of people of working age in 2020, which is similar to other Central and Southern Asian economies (Figure 4.13). The proportion is also set to expand further to 64.9% by 2035 before declining back to 63.1% in 2050. This trend contrasts with Europe and North America, where the working age population is contracting – 64.5% of the population in Europe and North America are of working age in 2020 but this is expected to decrease steadily to 61.2% by 2035 and to 58.7% by 2050 (Figure 4.13).

The old-age dependency rate is low. The proportion of the population aged over 65 in Kazakhstan is low at 7.9% in 2020, below developed countries in Europe and North America (18.3%), but above Central and Southern Asian countries (6.1%). Despite its low current level, the old-age dependency rate is expected to increase slowly to 9.3% in 2025, 11.1% in 2030 and 11.9% in 2035 (same data source as Figure 4.13). Overall, Kazakhstan does not face ageing related challenges at present.

Figure 4.13. Kazakhstan has a large and expanding working age population

Working age population (aged 15 – 64) projections in Kazakhstan and elsewhere, 2020 - 2050

4.4.2. Social security rates are low but rising

SSC rates are low by international standards. The SSC mix relies more heavily on employer rather than employee SSCs in Kazakhstan, which is common in most other OECD countries. Employer and employee SSC rates and payroll taxes are 14.5% and 10% in Kazakhstan in 2019. Taken together, these are currently below the OECD average (17.8% for employers and 9.8% for employees).

The current SSC rates are to be increased considerably, particularly for employers. For employers, the employer SSC rates for the existing insurance SSC and health SSC will be increased to 5% and 2% by 2020 respectively. Furthermore, the payroll tax will increase from 9.5% to 11% in 2025. In the past, employer payroll tax and health SSCs have been set with consideration given to each other. In 2018 for example, the social tax was reduced from 11.0% to 9.5% due to the introduction of the employer health SSC of 1.5%. In addition, the employer health SSC will be increased by a further 1% to 3% in 2022. The employee pension SSC will remain unchanged.

A new set of SSCs are to be introduced for employees and employers. A new employee health SSC, which will be withheld and paid by the employer, will be introduced at a rate of 1% starting from 2020 and then increasing to 2% in 2021. In addition, a new employer pension SSC will also be introduced in 2023 at a rate of 5%. The pension SSC was previously due to be introduced in 2020 but was postponed in 2019 due to concerns of the authorities related to the rising number of SSCs.

Box 4.3. The current SSC (and payroll tax) system

There are a number of SSCs and a social tax in Kazakhstan.

1. Employer payroll tax, also known as the social tax, is taxed at a rate of 9.5%. Taxes on payroll are generally defined as taxes paid as a proportion of payroll that do not give entitlement to social benefits. Payroll taxes are relatively uncommon in OECD countries. The employer payroll tax in Kazakhstan, also known as the social tax, is paid by employers on behalf of employees at a flat 9.5% rate in 2019. The rate is set to increase to 11% in 2025. As part of the 2007/2008 tax reform, a regressive payroll tax schedule at the time was replaced with a flat 11% rate. The rate was then reduced to 9.5% in 2019 in conjunction with the introduction of the new employer health SSC described below. Employer insurance SSCs (described below) are fully deductible from the payroll tax.

2. Employer health SSCs is tax at a rate of 1.5%. The employer health SSC, referred to as Obligatory Social Medical Insurance (OSMS), has a rate of 1.5% in 2019. The rate will increase further to 2% and 3% in 2020 and 2022 respectively. Employers remit the amount of the health SSC to the Social Medical Insurance Fund on a monthly basis.

3. Employer insurance SSCs are taxed at a rate of 3.5%. Employer SSCs, referred to as Obligatory Social Insurance Payments (OSCI), have a current rate of 3.5% in 2019. Following announcements in the 2019 State of the Nation Address, they are to be increased to 5% in 2020.9 These SSCs apply to local workers, citizens of CIS member countries and foreign workers with a permanent residence. As a result, foreign nationals with no permanent residence in Kazakhstan are not required to join certain SSC plans such as the pension plan.

4. Employee pension SSCs are taxed at a rate of 10%. For employees, the obligatory pension contribution (OPC) is levied at a rate of 10% on gross salary in 2019. For employers, a new mandatory employer pension contribution of 5% (of the income of employees) is to be introduced from 2023.

Source: OECD analysis.

If the proposed SSC rates increases are enacted, they will move above the OECD average. While both employee and employer SSC rates will be increased, overall increases will be greater for employers (Table 4.2). Currently, the total SSC rates for employees and employers are 10% and 14.5% respectively in 2019. These will increase to 11% and 16.5% respectively in 2020 and then further to 12.0% and 24.0% by 2025. In the case of employer SSC rates, insurance SSCs are deductible and these are shown separately. Overall, if these proposed SSC rates are implemented, they would push the overall SSC rate in Kazakhstan from below the OECD average to above it. This could occur as follows. The current combined rates of employee and employer SSCs are 24.5%, which is below the OECD average of 27.5%. However, SSC rates are expected to increase to 27.5% in 2020, which is about the same as the OECD average. By 2025, SSC rates in Kazakhstan will surpass the OECD average based on the current proposals. An obligatory professional pension contribution for employers, which applies to workers who work in dangerous working conditions, is also due to be introduced in 2020. However, this is not included in the analysis since it does not apply to all workers.

Type	SSC Category	Name	Status	2019 rates	Proposed rates (year of introduction)	Tax base	Tax base formula	Maximum cap	Remitted to
Employer	Payroll tax	Social tax	Current	9.5%	11% (2025)	Gross income less pension contributions and insurance SSCs deductible	[(Gross Income + Fringe benefits – Employee Pension Contributions) x 9.5%] - Insurance SSC	None	State Revenue Committee
	Health SSC	Obligatory social medical insurance	Current	1.5%	2% (2020) & 3% (2022)	Gross income less pension contributions	(Gross Income + Fringe benefits – Employee Pension Contributions) x 1.5%	15 MMS	Social Medical Insurance Fund
	Insurance SSC	Obligatory social insurance payments	Current	3.5%	5% (2025)	Gross income less pension contributions	(Gross Income + Fringe benefits – Employee Pension Contributions) x 3.5%	10 MMS	State Social Insurance Fund
	Pension SSC	SSC pension fund	Proposed		5% (2023)	Gross income		50 MMS	State Pension Fund
Employee	Health SSC	Obligatory social medical insurance	Proposed		1% (2020) & 2% (2021)	Gross income less pension contributions	(Gross Income + Fringe benefits – Employee Pension Contributions) x 1%	15 MMS	Social Medical Insurance Fund
Employee	Pension SSC	Obligatory pension contribution	Current	10%		Gross income	(Gross Income + Fringe benefits) x 10%	50 MMS	State Pension Fund

Notes: Pension contributions refer to obligatory pension contributions of employees. The obligatory professional pension contribution is not included as it applies only to a specific set of workers who work under dangerous working conditions. According to the Ministry of Healthcare, the State will make contributions for certain categories of socially vulnerable people including children, retired individuals and women on maternity and the contribution rates will be 4.0% in 2020, 4.5% in 2022 and 5% in 2025.

Source: OECD analysis; Ministry of Healthcare of Kazakhstan; IBFD; IMF (2020), Republic of Kazakhstan, Selected Issues.

Figure 4.14. The Government plans to increase overall SSC rates over the coming years

Employer and employee % SSC rates, 2019 - 2025

Figure 4.15. SSCs rates are scheduled to increase from below to significantly above the OECD average

Employee and employer SSC rates, % of gross income, 2018

	Employee SSC rates (%)			Employer SSC rates (%)
SSC type	2019	2020	2025	2019	2020	2025
Payroll tax				9.5	9.5	11.0
Insurance SSC				3.5	5.0	5.0
Health SSC		1.0	2.0	1.5	2.0	3.0
Pension SSC	10.0	10.0	10.0			5.0
Total rates1	10.0	11.0	12.0	14.5	16.5	24.0
Total rates (less employer insurance SSCs)	10.0	11.0	12.0	11.0	11.5	19.0

Notes: (1) While the total sums all of the employer SSC rates together, it should be noted that insurance SSCs are deductible from social tax. 2020 rates are based on announcements of authorities. Employer insurance SSCs and health SSCs are 5.0% and 2.0% from 2020 respectively. Employer pension SSC has been delayed to 5.0% from 2023. Employee health SSCs are 1% from 2020 and 2% from 2021. The obligatory professional pension contribution is not included in the total as it applies only to workers who work in dangerous working conditions (at a rate of 5%).

Source: Announcements of the authorities; IBFD; IMF, Republic of Kazakhstan, Selected Issues, February 2020.

4.4.3. The social security system is narrow by design

Some SSC contributions have ceiling caps. SSC ceiling caps, which place a limit on SSC contributions, are a policy design choice which are opted for by some but not all OECD countries. In Kazakhstan, maximum ceiling caps are applied to SSC contributions for insurance, health and pensions but not to payroll tax. The application of the maximum ceiling caps for pension SSCs and health SSCs are the same for employees and employers. The maximum gross income thresholds for SSC insurance (employer only), SSC health (employer and employee) and SSC pension (employer and employee) are 10, 15 and 50 MMS respectively. One rationale for the current set of SSC caps in Kazakhstan is the extent to which the social benefit is returned to the contributor in the future. This likelihood that the contributor receives the benefit at some point over the longer-term forms a basis for the divergence in the SSC caps. For example, pension SSC contributions are often returned to the contributor in the form of future pension payments, providing a basis for the higher maximum cap, whereas health SSC contributions are only returned when activated.

Review the deduction of employee pension contributions from the SSC base, which is atypical in OECD countries. In most OECD countries, employer SSCs are levied on gross income. In Kazakhstan, the SSC base for calculating SSC contributions is gross income less obligatory employee pension contributions. This approach to deduct employee pension contributions from the SSC base may imply that the pension contributions are seen as deferred income. If deferred pension income is not taxed today under employer SSCs, it would imply that a rationale exists to tax pensions upon distribution with the same employer SSCs that are levied on non-deferred wage income, such as health employer and employee SSCs.

Despite the social tax having a flat rate, the effective rate becomes progressive for higher incomes due to the deductibility of insurance SSCs, which have a maximum cap. Payroll taxes are typically levied at a flat rate on wages, which produces a constant average tax burden. As part of the previous 2007/2008 tax reform in Kazakhstan, the regressive social tax (or payroll tax) schedule at the time was replaced with a flat rate. Currently, insurance SSCs in Kazakhstan are deductible from social tax and therefore do not represent an additional burden on the employer (i.e. total employer social tax payable = social tax due – SSC insurance contributions). Since social tax and SSC insurance contributions are levied on the same tax base (i.e. employee gross salary plus fringe benefits less pension contributions) at 9.5% and 3.5% respectively, the overall effective social tax payable for employers is 6.0% (i.e. 9.5% - 3.5%). Furthermore, since social tax is not capped but SSC insurance is capped (at 10 MMS), the overall effective social tax payable will mechanically increase at higher incomes because the absolute amount of the deductible insurance SSC remains fixed above the cap and therefore becomes smaller relative to rising income. Consequently, the cap in effect produces a progressive social tax schedule (Figure 4.16).

Figure 4.16. Despite having a flat rate, the social tax becomes progressive at higher incomes

Payroll statutory and effective rates, for different income levels

4.4.4. The 2020 social security reform will increase the financing of the SSC system but further measures need to be considered

The inheritance of pension payments is generous, rare and expensive. One atypical policy design in Kazakhstan is that pension entitlements can be inherited, which is rare by international standards. It will require increasing SSC rates in order to finance this provision. Consideration could be given to abolish the inheritance of pension entitlements.

Employment costs for firms will rise over the coming years. Figure 4.17 shows the decomposition of the total employment cost for firms in 2019 and 2025 in Kazakhstan for single taxpayers without children that earn the mean gross wage in Kazakhstan into two main components: an individual’s take-home pay from labour (net of PIT and SSCs) and the tax wedge. The tax wedge is broken down further into its components, which consist of PIT and SSCs. On the basis of the analysis, employees have a tax wedge of 23.9% in 2019, a take-home pay of 76.1% and a total employment cost of KZT 179 877 (employment cost is gross mean wage plus social tax and employer SSCs). This means that almost one quarter of the total employment cost of an employee that earns the mean wage in the country goes to government as a result of taxes and SSCs, while just over three-quarters goes to the individual worker as their take-home pay. As evidenced in Figure 4.17, the category that contributes the most to the tax wedge in 2019 is individual SSCs (i.e. mainly pension employee SSCs) (KZT 16 367), followed by PIT and payroll tax. As discussed in section 4.4.2, between 2019 and 2025, a number of new SSCs will be introduced and current SSC rates will increase. For example, employer SSCs rates (excluding social tax) will increase from 5.0% to 13.0% and individual SSCs rates will rise from 10% to 11.8%. Since the insurance SSC (3.5% in 2019 and 5.0% in 2025) is deducted from the social tax (9.5% in 2019 and 11.0% in 2025), the effective social tax will remain the same at 6.0% of the tax base. In 2025, due to the introduction of the new SSCs and increases in the rates of SSCs existing in 2019, the tax wedge will rise to 30% and the total employment cost to KZT 195 572. Whereas the category of employer SSCs was a small contributor to the tax wedge in 2019, this category is substantial in 2025 (KZT 19 541 as opposed to KZT 7 365). Overall, employment costs are expected to increase by 8.7% based on the analysis. While increases in employee SSC rates are modest and may therefore not have a strong impact on labour market participation rates, the significant increase in employer SSC rate increases could reduce job creation in the private sector.

Figure 4.17. The tax wedge and total employment costs will rise in the coming years

Decomposition of total employment cost, employees in 2019 and 2025

The tax burden is flat across income levels. Figure 4.18 extends the analysis in Figure 4.17 by showing the tax wedge for employees in 2019 and 2025 for different levels of gross income. The average effective PIT rate and the personal average tax rates (which takes into account the PIT and employee SSCs as a share of gross wage) in 2019 are also presented. The analysis begins from about KZT 45,000 and the sharp increase in the average effective rate at KZT 63,125 is produced by the withdrawal of the 90% taxable income exemption, as discussed previously in (Figure 4.4). The analysis shows that the tax burden is relatively flat across different income levels, in particular once the strong progressive impact of the basic tax allowance is less dominant at higher incomes and because of the flat PIT and SSC rates. The analysis also highlights how the PIT rate burden is relatively low but the overall tax burden is high once SSCs are included.

Figure 4.18. The tax burden is flat across income levels

Decomposition of total employment cost, by gross income level (KZT), employees in 2019 and 2025

Frequent SSC policy changes coupled with delays increases administrative complexity and undermines economic activity. In recent years, there have been frequent and significant SSC proposed policy changes such as the introduction of new SSCs and changing current SSC rates (see Table 4.2). The increased number of SSCs to be calculated, remitted and paid by employers increases administrative complexity and the costs of doing business. Furthermore, some of these proposed policies have not be enacted or delayed. For example, the new employer pension SSC was previously due to be introduced in 2020 but was postponed until 2023 due to a concern of the authorities surrounding the rising number of SSCs. In addition, the SSC health contributions for individuals and the self-employed was delayed from 2018 to 2020. This combination of frequent proposed policy changes coupled with delays makes it challenging for individuals and businesses to plan for the future and undermines economic activity more generally.

The schemes that pay for the healthcare rely on a mix of different financing sources in OECD countries. Health systems can be financed through SSCs, general taxation or a combination of both sources of financing. Different approaches bring both advantages and disadvantages, as shown in Table 4.4. Government schemes typically receive budget allocations out of the overall government revenues. Social health insurance is usually financed out of social contributions payable by employees and employers. These schemes may also receive a varying proportion of their revenues from governmental transfers. Some countries are planning to reduce their reliance on wage-based contributions in the face of shrinking labour markets and financial shocks, and are increasingly looking for ways to diversify their revenue base (OECD, 2017[15]) (OECD, 2015[16]).

	Pros	Cons
General taxation	- Pool risks for whole population - Potential for administrative efficiency and cost control - Redistributes between high and low risk and high- and low- income groups in the covered population	- Risk of unstable funding and often underfunding due to competing public expenditure - Inefficient due to lack of incentives and effective public supervision
Health SSCs	- Generate stable revenues - Often strong support from population - Provides access to a broad package of services - Involvement of social partners - Redistributes between high and low risk and high- and low- income groups in the covered population	- Poor are excluded unless subsidized - Payroll contributions can reduce competitiveness and lead to higher unemployment - Complex to manage governance and accountability can be problematic Can lead to cost escalation unless effective contracting mechanisms are in place

Pros

Cons

General taxation

- Pool risks for whole population

- Potential for administrative efficiency and cost control

- Redistributes between high and low risk and high- and low- income groups in the covered population

- Risk of unstable funding and often underfunding due to competing public expenditure

- Inefficient due to lack of incentives and effective public supervision

Health SSCs

- Generate stable revenues

- Often strong support from population

- Provides access to a broad package of services - Involvement of social partners

- Redistributes between high and low risk and high- and low- income groups in the covered population

- Poor are excluded unless subsidized

- Payroll contributions can reduce competitiveness and lead to higher unemployment

- Complex to manage governance and accountability can be problematic

Can lead to cost escalation unless effective contracting mechanisms are in place

Source: International Labour Organisation. Adapted from (OECD, 2018[17]).

Arguments exist for financing social benefits through general taxation. Financing health insurance through a progressive PIT system could reduce labour and increase employment for low-income employees. There are some arguments to support the idea of financing social benefits through general taxation. For example, financing social benefits through general taxation, instead of SSCs, can reduce the tax burden on labour income through lower employer and employee SSCs that provide greater incentives for employers to hire workers and for workers to participate in the labour market. In the case of Kazakhstan, replacing flat SSCs with a progressive PIT also has the potential to enhance progressivity. The arguments for financing social benefits through general taxation are stronger when there is no strong link between the contributions made and the benefits received (Brys et al., 2016[18]). In general, when the contribution-benefit link is not strong, as is the case with health insurance in Kazakhstan, financing social security through a progressive PIT (or using taxes that bear not only on labour but also on capital income) could help reduce labour costs at low-income levels and increase employment, while also ensuring the financing of social security systems. On the other hand, benefits for retirement, disability and unemployment, which tend to be more strongly related to earnings, could remain financed in large part through SSCs. On the other hand, there are also limitations to fully shifting the financing of social protection from SSCs to PIT. The PIT base is narrow in Kazakhstan and has been further reduced by recent tax reforms to reduce the burden on low-income workers, which might not allow financing the same level of social protection. Generally, SSCs offer higher levels of social protection and give taxpayers an incentive to pay into the system. In addition, in a changing world of work, financing social benefits partly through general taxation could also ensure that welfare support remains available for a large number of people. Indeed, structural changes in the economy including digitalisation and automation are resulting in an increasing number of workers which make smaller SSC contributes (self-employed, temporary workers and workers with irregular working hours). These structural changes in the economy present new sustainability challenges for welfare systems that are financed primarily through SSC contributions.

4.5. Improve the design and functioning of the VAT

The design and functioning of the VAT can significantly be improved. The VAT in Kazakhstan is based upon the core features of a well-designed VAT, but there is scope to improve its design and administration. Most importantly, there is scope to broaden the VAT base by removing costly, ineffective and distortive preferential regimes. Probably the most important area where the VAT base can be broadened is the special VAT treatment for Special Economic Zones (SEZ). The current VAT treatment of SEZs constitutes a major flaw in the design of the VAT system and recently announced measures will not address the main design weaknesses. Instead, the authorities should fully restore the VAT chain by applying the standard VAT on all transactions to and within the SEZs while simultaneously providing more timely VAT refunds. Other areas for VAT base broadening include newly constructed residential buildings that are brought on the market for the first time and online sales of digital services and goods. Kazakhstan is currently in the process of adapting its VAT rules to bring inbound digital services and goods within the scope of the VAT. This reform will broaden the VAT base in line with the OECD International VAT/GST Guidelines. Significant scope also exists to improve the quality and capacity of the VAT administration. Although VAT revenue as a share of tax revenue is similar to the OECD average, the revenue performance of the VAT, measured by the ‘VAT productivity’ indicator, is relatively low, which points not only at a narrow VAT base but also at a weak tax enforcement and administration. Important steps have been taken by the authorities to address the challenges of the VAT refund system, and these efforts should continue. The VAT registration threshold in Kazakhstan remains high internationally, and continuing efforts to improve the VAT administration, to bring more firms within the formal economy and to simplify the VAT system should eventually allow Kazakhstan to lower the VAT registration threshold. Finally, there is scope to increase the standard VAT rate, which is currently low internationally. However, VAT base broadening measures and measures to increase VAT compliance should be implemented first before considering increasing the VAT rate. Moreover, any increase in the VAT rate should be considered in the context of inflation levels in the country.

VAT as a share of tax revenues are in line with OECD countries. VAT as a share of total tax revenues are similar to the OECD average and below the CIS average. VAT to total revenues is 19.2% in Kazakhstan compared to 20.2% in the OECD in 2017. Two-thirds of VAT relates to VAT on imports (68%) rather than domestic VAT (32%). However, VAT to GDP is lower than the OECD average (3.10% vs 6.8% in 2017) in part due to low overall taxes to GDP in Kazakhstan. VAT also comprises a relatively small 38% of the total taxes on goods and services. Most taxes on goods and services relate to taxes on specific goods and services for example on excises and customs duties, much of which comes from the natural resource sector. As a result, tax revenues on goods and services as a share of total taxes are high in Kazakhstan internationally at 51% compared to 32% in the OECD.

Figure 4.19. Taxes on VAT as a share of total taxes are in line with the OECD

VAT as % of total tax revenues, selected countries including from OECD and CIS, 2017

The VAT in Kazakhstan is based upon the core features of a well-designed VAT, but there is scope to further improve its design and administration. In a not so distant past, the authorities had considered reintroducing a sales tax similar to the tax that was levied before the introduction of the VAT in 1992. However, these plans have not been pursued, which has been the right decision, as sales taxes are distortive in contrast to a well-designed VAT. Instead, Kazakhstan has continued strengthening the design and operation of its VAT, following international best practice. This includes, for instance, the implementation of the reverse-charge collection method for cross-border B2B supplies of services. In light of the key role that the VAT plays in the total tax take in Kazakhstan, these efforts should continue including through further VAT base broadening and improved VAT collection mechanisms including for non-resident digital suppliers (see section 4.5.7 below).

VAT revenues have been increasing over time, largely aligned with growth and the general increase in price levels. The largest share of VAT revenues comes from imported goods and services although in recent years domestic VAT revenues are growing faster than VAT on imports.

	2011	2012	2013	2014	2015	2016	2017	2018
1. Net VAT revenues	865 087	914 194	1 327 433	1 197 258	943 051	1,495 682	1 664,699	2 034 314
Domestic	257 237	190 890	508 304	407 956	275 646	634 881	646 823	909 355
Imports	607 850	723 304	819 129	789 302	667 404	860 801	1 017 876	1 124 959
2. Refunds	306 000	451 000	249 000	n.a.	n.a.	498 378	612 084	505 011
3. Gross VAT revenues (1+2)	1 171 087	1 365 194	1 576 433	n.a.	n.a.	1 994 060	2 276 783	2 539 325
4. Non-oil and gas GDP	18 573 940	20 515 382	25 352 092	28 726 497	31 644 167	35 839 381	41 348 586	42 865 786
5. Net VAT collections as % of non-oil GDP (1/4)	4.66	4.46	5.24	4.17	2.98	4.17	4.03	4.75
6. Standard VAT rate (%)	12	12	12	12	12	12	12	12
7. VAT productivity (5/6)	0.39	0.37	0.44	0.35	0.25	0.35	0.34	0.40
8. Refunds as % of gross VAT revenues in % (2/3)	26.1	33.0	15.8	n.a.	n.a.	25.1	26.9	19.9

Notes: The domestic revenues reflect the sum of the categories “domestic VAT” and “other VAT” revenues as reported in the OECD Revenue Statistics database. Non-oil and gas GDP refers to the Gross Value Added (GVA) of the non-oil and gas sector as a share of GDP.

Source: State Revenue Committee of Kazakhstan; OECD Revenue Statistics database.

The revenue performance of the VAT in Kazakhstan, as measured by the ‘VAT productivity’ indicator, is relatively low. VAT revenue performance is defined as actual VAT revenues over non-oil and gas GDP divided by the standard VAT rate. It indicates the VAT revenues as a percentage of non-oil and gas GDP that is collected by each percentage point of the standard VAT rate (i.e. the share of the VAT-to-GDP ratio that is collected by each percentage point of the 12% standard rate). The VAT productivity ranged between 0.25 and 0.44 over the 2011-2018 period, with an average value of 0.36. This result is aligned with the average VAT productivity of 0.37 over the 2005-2013 period that has been found in (World Bank, 2015[19]) (the World Bank analysis applied data for non-oil GDP, while our analysis also excludes the gas sector from the analysis). The VAT productivity varies significantly over time but no clear (neither increasing or decreasing) trend can be observed. The VAT productivity measure represents an initial indicator and more work needs to be done to analyse the VAT compliance and policy gaps through a VAT tax gap analysis.

Box 4.4. The main features of a VAT

Although there is a wide diversity in the way VAT systems are implemented, VAT can be defined by its purpose and its specific tax collection mechanism. The OECD International VAT/GST Guidelines (2015) provide an overview of the core features of VAT, which are summarised below.

A broad-based tax on final consumption

A VAT is a tax on final consumption by households as, in principle, only private individuals, as distinguished from businesses, engage in the consumption at which a VAT is targeted. A necessary consequence of the fundamental proposition that a VAT is a tax on final consumption by households is that the burden of the VAT should not rest on businesses.

As a broad-based tax, the VAT is distinguishable from excises targeted at specific forms of consumption such as the purchase of gasoline or alcohol.

The staged collection process

VAT is collected by businesses through a staged process on the “value added” at each stage of production and distribution. Each business in the supply chain takes part in the process of controlling and collecting the tax, remitting the proportion of tax corresponding to its margin i.e. on the difference between the VAT imposed on its taxed inputs and the VAT imposed on its taxed outputs (see Figure below).

In general, jurisdictions with a VAT allow the deduction of VAT on purchases by all but the final consumer. This design feature gives to the VAT its essential character in domestic trade as an economically neutral tax. The full right to deduct input tax through the supply chain, except by the final consumer, ensures the neutrality of the tax, whatever the nature of the product, the structure of the distribution chain, and the means used for its delivery (e.g. retail stores, physical delivery, Internet downloads). As a result of the staged payment system, VAT thereby “flows through the businesses” to tax supplies made to final consumers. In practice, however, the right to deduct input tax may be restricted in a number of ways. Some are deliberate and some result from imperfect administration.

Infographic 4.1. VAT’s staged collection process – domestic supplies

The destination principle in international trade

According to the destination principle, the VAT taxing rights on cross-border supplies are to be allocated to the jurisdiction where the use or final consumption occurs. For cross-border supplies of goods, the tax is collected where the goods are imported. For supplies of services and intangibles, according to the OECD International VAT/GST Guidelines, taxation should occur where the business customer has located its permanent business presence for business-to-business supplies. For business-to-consumer supplies, the Guidelines recommend that the taxing rights over “on-the-spot supplies” be allocated to the jurisdiction in which the supply is physically performed; and that the taxing rights over all other supplies and services be allocated to the jurisdiction in which the customer has its usual residence. These include remote supplies of services and digital products over the Internet (e.g. apps, streaming of music and movies, online gaming) by foreign suppliers. The Guidelines recommend that these foreign suppliers be required to register and remit VAT in the jurisdiction of taxation and that countries implement a simplified registration and compliance regime to facilitate compliance for non-resident suppliers.

The relatively low VAT productivity (see Table 5.4) points at relatively high tax fraud and non-compliance and, linked to that, weak tax enforcement and administration as well as a narrow VAT base mainly because of the wide scope of VAT exemptions. Moreover, the fact that the VAT productivity indicator is not significantly increasing over time indicates that efforts should continue to strengthen the design (broadening the base and further lowering the VAT threshold) and administration of the VAT. Nevertheless, the VAT productivity indicator has jumped up from 0.34 in 2017 to 0.4 in 2018; data for 2019 will have to confirm whether the result for 2018 is an outlier or whether it may reflect a more permanent increase in VAT productivity. That being said, the VAT productivity indicator is a crude indicator that does not allow inferring the exact underlying drivers, and their changes over time, for the relatively weak VAT performance, so its results should be interpreted with care.

VAT revenues are predominantly paid by a few sectors and in particular the wholesale and retail sector. Other sectors, such as agriculture, accommodation and food and real estate activities contribute relatively little. Over half (56%) of all VAT relates to VAT on imports in Kazakhstan. However, the ratio is higher in some sectors. For example, in the wholesale and retail trade sector, which contributes more than any other sector to overall VAT revenues (44% of VAT in total), 78% of VAT relates to VAT on imports. Hence, a significant share of VAT revenues in Kazakhstan are paid by importing wholesale and retail companies that subsequently sell into the Kazakh economy (Figure 4.20).

Figure 4.20. Most VAT comes from the wholesale and retail sector and it is mostly VAT on imports

Gross VAT, domestic and imports, by sector in Kazakhstan, 2018

4.5.1. VAT rates

The VAT rate is low. The standard VAT rate of 12% is low compared to the CIS average of 17% and the OECD average of 19.3% in 2019. The rate is the same as in Kyrgyzstan but lower than the 15% applied in Uzbekistan and Turkmenistan, 18% applied in Azerbaijan and Tajikistan and the 20% rate applied in Ukraine, Moldova, Armenia, Belarus and Russia. In the case of Russia, an important tax policy change was undertaken in 2019 as part of the Russian Tax Code to increase the general VAT rate from 18% to 20%.

Figure 4.21. The VAT rate is low compared to the CIS and OECD average

Standard VAT rate in Kazakhstan, CIS countries and OECD countries, 2020

There are no reduced VAT rates but certain goods and services are zero-rated. A number of transactions are subject to a zero rate including exports of goods and services, international transport services, the sale of goods to businesses located in Special Economic Zones (SEZs) and amongst businesses within the SEZs, and the sale of goods produced in Kazakhstan to subsurface users that have the right to VAT-exempt importation of goods.

There is scope to increase the standard VAT rate in Kazakhstan in order to raise more revenues to finance the non-oil deficit. Until 2006, the standard VAT rate was 15 per cent, after which it was reduced by one percentage point each year, and has been stable at 12 per cent as from 2009 onwards. As part of a tax reform strategy that would narrow the non-oil deficit, government could consider increasing the VAT rate back to 15 per cent. For an average VAT productivity of 0.36 (see above), an increase in the standard VAT rate by 3 percentage points would raise additional revenues of 1.1% of GDP (under the assumption that such a change would not result in large behavioural changes). However, as tax distortions are typically increasing in the tax rate, VAT base broadening measures and measures to increase VAT compliance should be implemented first before considering increasing the VAT rate. Indeed, VAT rate increases might just exacerbate the distortions that result from the narrow base and weak compliance, so it is important that efforts are taken in these areas first. Otherwise, a VAT rate increase might just further increase the informal economy without leading to more tax revenues. Policymakers might be tempted to introduce a reduced VAT rate when they increase the standard VAT rate, but such a reform would not be efficient nor fair.

A VAT rate increase should be considered in the context of an inflationary environment. Increasing the VAT rate is in effect a once-off increase on consumers. However, the inflation rate in Kazakhstan is high at 7.4% in 2017 compared to the CIS average of 5.7% and the OECD average of 1.8%. Furthermore, and as discussed in Chapter 2, there are a number of inflationary pressures in the Kazak economy including expansive government expenditure programmes and rising credit and government debt. While Kazakhstan may tempted to follow the recent example of Russia by increasing the VAT rate to raise revenues, it should be noted that inflation is considerably lower in Russia at 1.8%.10

4.5.2. VAT exemptions

Many goods and services continue to be exempt from VAT, and Kazakhstan has taken efforts to broaden its VAT base by bringing certain goods and services, such as gambling, within the VAT base. While all VAT regimes have VAT exemptions and VAT reduced rates are more the rule than the exception in the OECD, there remains scope in Kazakhstan to broaden its VAT base. The goods and services that are VAT-exempt in Kazakhstan include, amongst others, the sale of newly constructed residential buildings that are brought on the market for the first time, financial services (including banking and insurance), funeral services, social services provided by non-profit organisations and services provided in the cultural, science and education sector. Duty-free goods are also VAT exempt as well as goods and services related to medical and veterinary activities (including medicine and medical assistance), and certain services performed by tourism operators.

In general, the use of reduced VAT rates or exemptions should remain limited. As with any preferential tax treatment, reduced VAT rates and exemptions narrow the tax base and reduce potential revenues. Reduced VAT rates and exemptions are introduced for a wide range of purposes, but overall they have been found to be poorly targeted instruments to support low-income households, to support consumption or support certain business activities or sectors. This applies also for VAT reductions that apply to basic necessities. At best, rich households receive roughly as much benefit – in absolute value – from a reduced rate as do poor households. At worst, rich households benefit vastly more than poor households. This result is unsurprising as better off households consume more, and often more expensive, products than poorer households. Thus, while poorer households may benefit from reduced VAT rates on necessities, the wealthier gain even more (OECD/KIPF, 2014[20]). Targeting support at low-income households is often best achieved through the transfer system (e.g. direct cash transfers), particularly in countries with well‑developed social security systems.

VAT exemptions lead to cascading effects (except if applied in the final stage of supply changes) and can encourage the granting of further exemptions to prevent this issue. Unlike reduced VAT rates or a zero-rate where suppliers charge the VAT to their customers at a reduced or zero rate but keep the full right to deduct the VAT paid on their purchases, under the VAT exemption the suppliers do not charge any VAT to their customers but are not allowed to recover any VAT on their inputs. As a consequence, the input VAT becomes a cost for businesses selling VAT-exempt products. This input VAT is likely to be embedded in the price of exempt products. In fact, businesses might either shift that extra tax burden onto their customers by raising sale prices or bear (part of) the cost of unrecovered VAT themselves through a reduction of their profit margin. As a result, VAT exemptions distort the input choices of businesses that face an incentive to buy inputs with little or no tax on them (including inputs purchased from the informal sector). Exemptions also create the incentive for businesses to undertake activities they otherwise would outsource. Exemptions can also discourage investment, as sellers of VAT-exempt products will not recover the VAT paid on the purchases of new investment. In some countries, this has led to pressures to grant further VAT exemptions on the inputs used by suppliers of VAT-exempt products, but such a strategy just aggravates the distortions introduced by exempting certain transactions.

Kazakhstan should continue its efforts to broaden the VAT base. In line with previous analysis (World Bank, 2008[21]), there are a number of goods and services that could be brought within the scope of the VAT aligned with international VAT best practice. These include newly constructed residential buildings that are brought on the market for the first time and tourism services provided to foreigners and, possibly to a lesser extent, insurance-related services and cultural activities. Government is encouraged to evaluate each VAT exemption from an efficiency, equity and revenue perspective. Such an in-depth cost-benefit analysis would then provide objective arguments against or in favour VAT base broadening. Another VAT base broadening measure would be to continue lowering the VAT registration threshold (see section 4.5.8 later in the chapter), which currently remains high and may partly explain why the agriculture, accommodation and food and real estate activities contribute surprisingly little to the VAT revenues.

VAT base broadening might allow raising additional VAT revenues without having to increase VAT rates to finance the non-oil deficit. A moderate increase in the VAT productivity indicator from the average value of 0.36 (over the 2011-2018 period) (see Table 4.4) to 0.4 (i.e. an increase by 0.04), for instance, would yield additional VAT revenues equal to 0.48% of GDP (i.e. 0.04 times the standard 12% VAT rate) if the VAT rate would remain 12%. The VAT base broadening would yield higher additional tax revenues if, at the same time of the VAT base broadening, the standard VAT rate were to be increased.

4.5.3. The VAT refund system

A well-designed VAT refund system is an essential part of a well-functioning VAT system. The VAT is collected by businesses through a staged process on the “value added” at each stage of production and distribution. In general, a VAT registered business is allowed to offset the VAT paid on its inputs against the VAT collected from its customers, and remits the difference to tax authorities. In cases where input VAT exceeds output VAT, a VAT credit arises. In order to protect the revenue base against false VAT input claims, tax administrations typically audit businesses before making a payment of a VAT credit.

Kazakhstan follows best practice and pays VAT credits only after a tax audit. However, the maximum allowed refund period is quite long (155 calendar days) but businesses that export more than 70 per cent of their turnover receive a refund of excess VAT within 55 days. In addition, Kazakhstan implements a simplified VAT refund procedure that does not require a tax audit. This regime applies to businesses that are monitored as large taxpayers or businesses that fall under the horizontal monitoring program. In general, delays in VAT refunds imply an opportunity cost for businesses in terms of the time value of money and can generate significant cash-flow problems. It increases incentives for artificial integration of businesses in order to be able to offset tax credits on exports against VAT on domestic sales. It can also constitute a drag on investment and provides businesses with a tax-induced incentive to spread investment over time rather than investing more quickly. As a result, Kazakhstan could continue to reinforce its auditing activities such that a more timely VAT refund can be implemented for small and large firms, irrespective of whether they sell to the domestic or foreign markets.

The wide range of activities that are VAT zero-rated, in particular activities that take place in Special Economic Zones, increase the size of the VAT refund claims considerably. When a business sells zero-rated goods to businesses located in a SEZ, it can claim a refund for the VAT paid on its inputs, which prevents the distortions created by VAT exemptions, as discussed above. However, the zero-rating of transactions breaks the VAT chain, which will have a negative impact on VAT revenues (see below), but it also increases the challenges faced by the VAT refund system. Restoring the VAT chain by including zero-rated supplies for transactions with businesses in SEZs within the standard VAT system will therefore prevent too large VAT refund claims, which will also free-up the tax administration’s audit capacity.

While the challenges of the VAT refund system have been identified for many years as one of the areas of policy reform in Kazakhstan, important steps have been taken already. For instance, the time to obtain a VAT refund has recently been shortened to 30 days for businesses that issue and receive exclusively electronic VAT invoices for the tax period for which a refund is claimed. Overall, the recent introduction of electronic VAT invoicing in Kazakhstan is a step in the right direction and the efforts to streamline it across all businesses should continue.

Nevertheless, the VAT refund system can be further strengthened as, in practice, obtaining refunds continues to require significant efforts. This was acknowledged in the State of the Nation address 2019, which has highlighted the need to apply a simpler and faster VAT refund procedure. A possibility to help enhance the efficiency of VAT refunds even further could be to introduce a system that would identify low-risk businesses and refund them faster; this could be part of a strategy that implements a well-organised refund system based on clear design principles within a risk-based compliance management strategy.

In addition, Kazakhstan implements a number of limits to its VAT refund system, both in the amount of the refund and the timing of the refund (e.g. with respect to input VAT regarding transactions whose place of supply is not in Kazakhstan, VAT levied on imports, etc.) in order to protect its VAT base and avoid VAT fraud. Over time, when the VAT audit system is further strengthened, these limitations could be revisited and, possibly, changed. In addition to the analysis included in this report, there is merit to conduct an in-depth evaluation of the VAT refund system in Kazakhstan. This is left for future work.

The pressures on the VAT refund system should be tackled at source by improving the design of the VAT in Kazakhstan. Part of the pressures on the VAT refund system are a consequence of the flaws in the design of the VAT in Kazakhstan, including the VAT treatment in relation to SEZs. Improving the design of the VAT will then also alleviate some of the pressures on the VAT refund system. The level of refunds can be reduced through several measures, for instance by introducing a deferred payment system for imported capital goods combined with a reverse-charge mechanism which transfers the VAT liability from the import stage to the domestic buyer who is allowed a VAT input credit against any output VAT levied on its sales (World Bank, 2015[19]). Abolishing the VAT zero-rating of activities that take place in SEZs and any other VAT zero-rating for domestic transactions would be another reform that would reduce the pressure on the VAT refund system in Kazakhstan.

4.5.4. Special economic zones

The special VAT treatment for Special Economic Zones is a major flaw in the design of the VAT in Kazakhstan. Transactions from the domestic economy with businesses located in a SEZ, as well as transactions amongst businesses in a SEZ, are zero-rated. This break in the VAT chain puts pressure on the VAT refund system. As businesses in the domestic economy might face difficulties in obtaining a VAT refund, SEZ suppliers might simply add the 12% VAT rate, or part thereof, in the price they charge. Moreover, it creates opportunities for fraud, as goods and services sold by businesses located in a SEZ to the domestic economy might get untaxed. It creates incentives for domestic businesses to create a subsidiary in a SEZ to engage not only in profit shifting, but also to organise transactions such that the overall VAT liability is minimised (in light of the weaknesses in the VAT refund system). Moreover, it increases the risk that zero-rated inputs for businesses in a SEZ are diverted to the domestic market without the payment of the VAT. Because of these risks for VAT fraud, the tax administration is required to devote valuable resources to administer and enforce the VAT system despite the fact that the SEZs themselves are, to some extent, outside of the VAT system. Further strengthening the VAT refund system such that businesses located in a SEZ receive a VAT refund in a timely manner when they export their goods and services is a much better strategy than the current zero-rating.

Recently announced measures will not address the main weaknesses in the VAT design for SEZ businesses. As from April 2019, government allows suppliers of certain goods to businesses located in a SEZ to opt for the standard rather than the 0% rate. This option is available only if the goods are fully used by the activities carried out within the SEZ business. The reform measure will reduce the need to refund VAT for businesses that sell goods to businesses located in a SEZ as they can charge output VAT from which they can deduct their input VAT. However, the reform merely shifts the excess VAT credit to businesses from outside to businesses within the SEZ. SEZ businesses are VAT 0%-rated and therefore only charge VAT on their sales to the domestic market (and possibly to other businesses within the SEZ if they would qualify for the above measure). SEZ businesses that have to pay input VAT but do not charge output VAT will very likely need to be reimbursed for the input VAT paid.

This reform indicates that government is aware of the major challenges created by the 0%-rating in SEZs, but in order to address the challenges, more far-reaching measures are required. Although this measure creates an incentive for SEZ businesses to charge the standard VAT rate, the reform is only a partial solution that further increases complexity, tax uncertainty for businesses and tax administration costs. Instead, government should fully restore the VAT chain by applying the standard VAT rules to SEZs, which means levying the standard VAT rate on all transactions to and within the SEZs and providing a timely VAT refund, for instance when goods are exported.

The introduction of a deferred payment system for imported capital goods should also apply for businesses that are located in a SEZ. Such a reform will therefore not only reduce the pressure on the VAT refund system, but it would also take away one of the main arguments in favour of a preferential VAT treatment of capital-intensive businesses located in SEZs. It is important to keep in mind that such a deferred payment system is complex to operate in practice and extremely vulnerable to fraud. Its introduction would there require a well-prepared and efficient VAT administration as a prerequisite. Overall, further strengthening the VAT refund system, as discussed above, will prevent the need for a special VAT treatment for SEZs, which is a key reform priority.

Box 4.5. The distributional effects of reduced VAT rates in OECD countries

With the exception of Chile, all OECD countries have one or more reduced VAT rates to support various policy objectives. A major reason for the introduction of a differentiated rate structure is the promotion of equity. Countries have generally considered it desirable to alleviate the tax burden on goods and services that form a larger share of expenditure of the poorest households (e.g. basic food, water). Countries also often decide not to tax medicine, health services and housing at high rates. Reduced VAT rates have also been used to stimulate the consumption of “merit” goods (e.g. cultural products and education) and other non-distributional objectives such as promoting locally supplied labour-intensive activities (e.g. tourism) and correcting externalities (e.g. energy-saving appliances).

In general, VAT exemptions, zero-rates and reduced rates are not a well-targeted tool to support low-income households. Reduced rates that are implemented in countries for the distinct purpose of supporting the poor (i.e. to address distributional goals) typically do have the desired progressive effect. For example, reduced rates for basic food provide in general greater support to the poor than the rich as a proportion of household income or expenditure. However, despite this progressive effect, these reduced VAT rates are a very poor tool for targeting support to poor households. At best, rich households receive roughly as much benefit – in absolute value – from a reduced rate as do poor households. At worst, rich households benefit vastly more than poor households. This result is unsurprising as better off households can be expected to consume more, and often more expensive, products than poorer households. Thus, while poorer households may benefit from reduced VAT rates on “necessities” the wealthier gain even more.

Cash transfer programmes that cover the entire population, if well-functioning, are a more effective tool to compensate poor households for the VAT they have paid. If poor households can be compensated directly through a cash transfer programme, it is more efficient and fair to tax all goods and services at the standard VAT rate and compensate the poor directly through cash transfers (and/ or reductions in personal income taxes, etc.), especially if the standard VAT rate is not particularly high. It should immediately be noted, however, that compensating all (and only the) losers of a reform through a transfer programme might in practice be very difficult to achieve.

With regard to preferential VAT provisions for social, cultural and other non-distributional goals, richer households benefit considerably more from VAT exemptions and reduced rates. Those tax provisions often provide so large a benefit to rich households that the reduced VAT rate actually has a regressive effect – benefiting the rich more both in aggregate terms and as a proportion of expenditure. For example, reduced rates on hotel accommodation and restaurant food benefit the rich vastly more than the poor, both in aggregate and proportional terms, in all OECD countries in which they are applied. Similar results, but of less absolute magnitude, are found for reduced rates on books, cinema, theatre and concerts.

Finally, VAT rate differentiation might not be the best policy instrument to correct negative externalities. VAT rate differentiation may improve efficiency if it means that the private marginal costs of an activity are brought closer to the marginal costs for society. However, VAT is a blunt instrument for addressing environmental externalities, as it may be hard to target the actual source of pollution. For example, reduced rates on energy-saving appliances may boost demand or them and therefore stimulate the consumption of these goods. The reduced VAT rate may give incentives to shift from more to less energy-consuming items (consumers might replace their old refrigerator with a new one, for instance). However, this may also lead to an increase in the purchase of energy-intensive products (e.g. consumers may replace their old refrigerator with a new refrigerator and a freezer).

Source: OECD/KIPF (2014)

4.5.5. Cross-border trade in services and intangibles

Kazakhstan is in the process of adapting its VAT rules to increasing digitalisation and online sales, but the implementation has been delayed until 1 January 2021. Regarding cross-border trade in services and intangibles, Kazakhstan is in the process of reforming its VAT rules and practices to levy VAT on inbound digital supplies, such as software provided over the internet, online advertising, and online data storage, downloads of media and e-books, and website services. Foreign providers and digital marketplaces will be required to register, charge VAT and pay the amount to the tax administration in Kazakhstan.

This reform will broaden the VAT base by ensuring the taxation of inbound digital supplies, in line with the OECD International VAT/GST Guidelines. According to the OECD International VAT/GST Guidelines, taxing rights on cross-border business-to-consumer (B2C) supplies of services and intangibles (including digital supplies, e.g. apps, streaming of music and movies, online gaming) are to be allocated to the jurisdiction in which the customer has its usual residence. The OECD Guidelines recommend that these foreign suppliers be required to register and remit VAT in the jurisdiction of taxation and that countries implement a simplified registration and compliance regime to facilitate compliance for non-resident suppliers.

4.5.6. Cross-border sales of low-value goods

Regarding cross-border sales of goods, most countries have been operating VAT relief regimes for imports of low-value goods, as the costs of collecting VAT on those items through traditional customs processes were often likely to outweigh the VAT actually collected. At the time when most of these relief regimes were introduced, online shopping did not exist and the level of imports benefitting from the relief was relatively small. However, there has been a significant and rapid growth in the volume of imports of low-value goods subject to these VAT relief regimes. This has resulted in large potential VAT revenues not being collected and growing risks of unfair competition for domestic retailers that are required to charge VAT on their sales to domestic consumers.

Digitalisation is pushing governments to revise their VAT rules on cross-border trade in low-value goods. Countries that wish to collect VAT on low value imports, particularly imports from online sales by foreign vendors, could do so by increasing the efficiency of the processes of VAT collection on these imported items. The OECD has developed internationally agreed standards that allow countries to improve the VAT collection by moving the requirement to collect and remit the VAT away from the order (customs procedures) to the online vendor or an intermediary (such as a digital platform; see below). These standards were included in the BEPS Action 1 Report and in two subsequent OECD reports that were adopted by the Inclusive Framework and by the Global Forum on VAT as consensus solutions.

These standards are now gradually implemented across the OECD, which could inspire Kazakhstan to follow this best practice. Australia was the first OECD country to implement these standards, effective as of July 2018, and has already reported revenues amounting to AUD 81 million in the first quarter of operation of the regime. New Zealand has introduced a similar regime in 2019 and the European Union announced its introduction in 2021. Kazakhstan may wish to assess whether the current VAT relief regime for imports of low-value goods (or the import of all goods by individuals “not for entrepreneurial purposes”) poses issues. In particular, authorities could evaluate the extent to which the VAT relief regime puts domestic businesses at a competitive disadvantage and how much VAT revenue they forego because of that measure.

4.5.7. Digital platforms

Digital platforms have the potential to further increase the efficiency of VAT collection on online sales given that the majority of these sales occur through a relatively limited number of platforms. The OECD has started in 2017 to develop agreed standards on the potential role of platforms in facilitating and improving VAT collection and compliance. Involving digital platforms could imply, for instance, that if goods are purchased via an online marketplace, the online marketplace will be treated as the supplier of the goods and will therefore be responsible for collecting and remitting the VAT. A key reasoning behind this approach is that the platform is viewed as taking the role of a ‘store’ with an offering of different supplies and in many cases act as the sole point of contact with the end consumer ( (OECD, 2019[22])). Several jurisdictions have already introduced or announced the introduction of measures involving the digital platform in the collection of VAT on sales of digital services via platforms. In addition to setting the standards, the OECD provides practical guidance to tax authorities on how to make digital platforms liable for the VAT on sales made by online traders through them, along with other measures that include data sharing and enhanced cooperation between tax authorities and digital platforms (Box 4.6).

In addition to digital platforms involved in the supply of B2C cross-border supplies, online booking platforms could be involved in the collection of VAT on the sales that they facilitate. In this case, for example, the VAT could be levied on the price inclusive of the booking platform’s margin. Involving booking platforms in the collection of VAT would also help address informality. Under a system where digital platforms would be fully liable for collecting and remitting VAT, VAT would be imposed on all online sales, including those by informal suppliers operating through a platform. A threshold could be set at the platform level, under which no VAT would be levied. As mentioned above, digital platforms could also provide information to the tax administration, as third party reporting is a key way to strengthen tax compliance.

Box 4.6. OECD report on ‘The role of digital platforms in the collection of VAT/GST on online sales’

The OECD has provided guidance in the International VAT/GST Guidelines (the “Guidelines”) to jurisdictions wishing to collect VAT on cross-border supplies of services and intangibles. The Guidelines include recommended rules and mechanisms for the effective collection of VAT on business-to-consumer (B2C) supplies of services and intangibles (including digital supplies) by foreign suppliers. The Guidelines were complemented by the 2017 report on Mechanisms for the effective collection of VAT/GST where the supplier is not located in the jurisdiction of taxation and the 2019 report on The role of digital platforms in the collection of VAT/GST on online sales, which provide guidance on implementation to jurisdictions.

It was indeed recognised that platforms may significantly enhance the effectiveness of VAT/GST collection given their important role in generating, facilitating and/or executing online sales. In fact, a number of jurisdictions have already implemented measures to involve digital platforms in collecting VAT/GST on online sales and have reported positive outcomes in securing tax revenue. Other jurisdictions are considering the introduction of such measures. These reports of course do not aim to provide detailed prescriptions for national legislation, but rather seek to present a range of possible approaches as a reference point to assist policy makers.

4.5.8. Registration thresholds

In general, setting the VAT registration threshold at an adequate level is a complex task. The main reason for excluding small businesses from the VAT system is that compliance costs for small businesses may be disproportionate compared to their turnover, and that the costs for the tax administration of having very small businesses pay VAT may be disproportionate compared to potential VAT revenues. On the other hand, a VAT registration threshold introduces competitive distortions between small businesses under and above the threshold. Overall, the VAT registration threshold should minimise competitive distortions and be set so that the revenues collected are higher than the administrative costs of ensuring that small businesses properly collect and remit VAT. In general, a higher threshold is considered more appropriate in countries where the tax administration tends to be weaker.

The VAT registration threshold in Kazakhstan has been decreasing over time but remains high internationally. Registration as a VAT payer is required for all individuals and companies that conduct business activities in Kazakhstan and have cumulative taxable revenues in excess of MCI 30 000 in 2019 (~USD 200 000) in any consecutive 12-month period. Taxpayers whose annual turnover is less than this amount can voluntarily register for VAT (World Bank, 2015[23]).

The tax administration should strengthen its operation so that the VAT registration threshold can gradually be lowered over time. As the tax administration’s capacity becomes stronger, Kazakhstan could consider lowering the VAT exemption threshold. In addition, with the rise of the sharing economy and the possible increase in the number of small operators below the VAT registration threshold, the revenue loss and distortions caused by a high VAT registration threshold might become more problematic.

A lower VAT registration threshold could also be accompanied by VAT simplification measures. In Kazakhstan there are potentially significant administrative costs associated with registering for VAT including penalties for late issuance of electronic invoices and the exclusion of the amount of VAT from the credit if transactions with a noncompliant counterparty are revealed. As a result, lowering the VAT registration threshold should be accompanied with simplification measures. One possibility would be the introduction of a VAT flat rate scheme. A flat rate scheme allows eligible businesses (e.g. below a certain turnover threshold) to apply a fixed and lower VAT rate, typically to turnover, to determine VAT due (OECD, 2015b). In principle, under such schemes, businesses give up the right to reclaim VAT on their inputs. Businesses therefore keep the difference between what they have charged their customers and what they pay to the tax administration. Variable flat rates may be applied, and are intended in most cases to reflect the average VAT rate in specific industries or sectors after taking into account the recovery of VAT on inputs. From a tax compliance perspective, a major advantage of these schemes for small businesses is that they are not required to keep detailed records of sales and invoices. However, creating a special regime that may over-compensate small businesses provides new opportunities for tax avoidance and evasion and may reduce firms’ incentives to grow. Alternative measures may be preferred to simplify compliance and administration, including for instance less frequent filing of VAT returns and simplified accounting methods.

Figure 4.22. The VAT registration threshold is high

Compulsory VAT registration thresholds in US dollars

4.5.9. Strengthening VAT compliance

VAT fraud takes many forms. Common forms of VAT fraud involve businesses that should be registered for VAT not registering, by remaining completely informal, artificially splitting activities into smaller businesses or under-reporting sales to remain under the compulsory VAT registration threshold. Non-compliance may also occur with VAT-registered businesses: some may for instance under-report taxable supplies (e.g. through automated sales suppression devices or “zappers”) or overstate purchases for which they can deduct input VAT (through false invoices); others may even disappear without remitting VAT to the government.

There are available tools to deal with sales suppression. At a basic level, sales suppression can be as simple as not recording some cash sales with the intention of under-reporting the amount of sales and thereby under-reporting the corresponding tax liability. More sophisticated methods have also become prevalent, with sales suppression being undertaken through electronic tools that can alter evidence of transactions whether paid in cash or card, without leaving a trace of the alteration. The most common counter-suppression tool used to address electronic sales suppression is data recording technology. This tool records and secures the sales data immediately as the transaction occurs and stores it in a manner that means it is tamper proof. There are different types of tools that are being used to perform this function, which are referred to in different countries and by different service providers as a fiscal control unit, an electronic fiscal device, or sales data controller. This type of technology should be applicable to any type of cash register, such as traditional electronic cash registers (ECRs), computer-based point of sales systems, or those that are tablet or smartphone-based. These tools are also being used to send data automatically to the tax authority, connecting cash registers online to their data server systems. This can occur either in real time or in bulk scheduled transfers, e.g. at the end of the day or each month. The tax authority then has the opportunity to access the data remotely for compliance and audit purposes. Results from these devices have been impressive. For instance, in Hungary, electronic cash registers were installed with a fiscal control unit. After the first year of operation, VAT revenue increased by 15% in the targeted sectors (OECD, 2017[24]).

Kazakhstan has introduced the e-system of invoicing aligned with international best practice. There are other tools to deal with businesses falsifying invoices. Whereas sales suppression techniques seek to under-report revenue, false invoicing seeks to over-report deductions, and to falsify invoices to mask non-deductible personal expenses as legitimate deductions. A solution to address the problem of false invoicing is requiring electronic invoicing. Generally, businesses must retain records of transactions with customers and provide an invoice to a customer, either in electronic or paper form. An electronic invoice documents the transaction in electronic format. The electronic invoicing system should have additional features to ensure the integrity of the data and the identity of the creator. This can be done by using a digital signature to ensure the authenticity of the invoice and that it has not been altered after its creation. Electronic invoicing will be most effective where the invoices must be registered or otherwise provided to the tax authority. The detection of false over-reporting of deductible expenses can be achieved by automatically matching the data for the purchaser and seller. Where this is undertaken through periodic or real time data transfers, the tax authority has substantially enhanced visibility of its taxpayers, and can perform audits, analytics and tax return functions in an efficient way. Electronic invoicing has been implemented in a number of countries, with evidence beginning to be collected on its impact.

Kazakhstan should continue to develop its risk management system. Risk-based approaches enable tax authorities to better identify high-risk businesses and fraudulent behaviours and therefore contribute to targeting limited government resources more effectively. The risk-based analysis system can be based on crossing digitalised data collected by tax authorities with data from other sources, including data from the customs administration, data from real estate and vehicle registers, data from different business databases, etc. In this context, machine learning can also be a powerful tool to uncover hidden patterns in the existing data. Kazakhstan introduced a ‘Risk Management System’ in 2018 using data mining, mathematical modelling and predictive analytics, inter alia. The system has recovered an additional KZT 47 billion and KZT 98 billion in 2018 and 2019 respectively. (Division of Automation of Business Processes, State Revenue Committee).

4.6. Other indirect taxes

4.6.1. Revenues from excise duties support the financing of the healthcare system

Excise taxes are low compared to OECD countries and could be an area of future analysis and reform. Excise taxes apply to the sale and import of crude oil, gas, petrol, diesel, alcohol, tobacco and passenger cars (exports are not subject to excise taxes; export customs duties on oil relate to oil products and are not included for this analysis). For domestically produced and sold excisable goods, the tax base is volume. Excise taxes are levied on the pre-VAT selling price. Excise taxes represented 2.9% of total tax revenues in 2017 in Kazakhstan, significantly less than the OECD average of 7.6%. In 1998, the share of excise taxes in total revenues was similar to the OECD average but has fallen significantly as the economy grew over recent decades. Excise taxes are mostly comprised of tobacco (54%), alcohol (23%) and petroleum products (23%). Excise taxes could be an area for future analysis and reform in Kazakhstan in part because excises are relatively straightforward to administer and most excisable goods are inelastic.

Arguments exist to finance public healthcare partly with excise duties on alcohol and tobacco. Most excise tax revenues in Kazakhstan come from alcohol and tobacco in 2017 (77%). Excise duties on these goods internalise some of the costs that their consumption imposes on society. For example, the consumption of alcohol and cigarettes increase healthcare costs directly while also indirectly reducing labour market productivity through sick leave. At the same time, improving Kazakhstan’s healthcare system will require more financial resources (discussed in section 4.4). Consequently, there may be a case for using some portion of excise duties to support healthcare, as is currently done in some OECD countries (OECD, 2015[16]). On the other hand, arguments also exist against the earmarking of tax revenues. Earmarking can reduce flexibility in government budgeting if, for example, more revenue is raised than necessary in a particular year and the excess revenue cannot be used elsewhere to support budget shortfalls.

Figure 4.23. Excise tax revenues are low

Excise taxes as a share of total tax revenues, Kazakhstan and the OECD, 1998 - 2017

4.6.2. Strengthen the role of recurrent taxes on residential property

Property and land taxes play a minor role in Kazakhstan. Recurrent taxes on business immovable property are paid each year at low and progressive tax rates, ranging from 0.1% to 1.5% depending on the taxpayers activities. Recurrent taxes on business property constitute the largest share of all property taxes paid in Kazakhstan. Recurrent taxes on household immovable property are, however, very low. In addition, income received by an individual (who is not an entrepreneur) from renting out immovable property is treated as business income. The capital gains realised from the disposal of immovable property (other than a taxpayer’s permanent residence) are taxable. Taxpayers that own land must also pay land tax each year at varying rates (depending on the use, size and quality of the land). Overall, property tax revenues are low, representing 3.0% of total tax revenues in 2017, compared to the OECD average of 5.8% (Figure 4.24). The share of property tax in total tax revenues has remained stable in recent years in both Kazakhstan and the OECD. Land taxes as a share of total revenues were very low at 0.17% of total tax revenues in 2017.

Almost all property taxes in Kazakhstan are paid by companies instead of households. According to data from the State Revenue Committee, 96% of property tax revenue is paid by 135,900 companies in 2018. The remaining 4% of property tax revenue is paid by 3.15 million households. This is also reflected in OECD revenue statistics data, which shows that household recurrent taxes on immovable property (4110) are very small in Kazakhstan at 0.05% of tax revenues. Furthermore, and as shown later in Table 5.2, taxes on property are highly concentrated with 99% of all property taxes paid by companies in the top turnover decile and 59% of all property taxes paid the top 100 companies in terms of turnover.

Recurrent taxes on residential property are an efficient form of taxation. In OECD countries, empirical analysis concluded that recurrent taxes on immovable property, in particular when owned by households, were the least damaging tax for long-run economic growth, compared to consumption taxes, personal income taxes and corporate income taxes (Johansson et al., 2008[25]). Recurrent taxes on immovable property can be efficient because the tax base – typically land and improvements – is immobile, which limits the behavioural response to the tax. The visible nature of immovable property also makes the tax harder to evade. If properly designed, recurrent taxes on immovable property can also be designed in a fair manner. This can be achieved by, for instance, exempting a certain amount of residential property wealth from tax.

Recurrent taxes on business property might discourage investment. While recurrent taxes on residential property might be the least harmful tax for economic growth, this might not hold for recurrent taxes on immovable property levied on business property. Depending on their design, these taxes might create an additional tax burden on investment, which is levied irrespective of the profits earned by the business. These taxes should be kept as low as possible. Transaction taxes are also distortive and their rates should be kept low, except perhaps in situations of housing price bubbles. Inheritance taxes, on the other hand, do have a role to play in a well-designed country tax system.

Kazakhstan should strengthen the role of recurrent taxes on residential property. Increasing the role of recurrent taxes on residential immovable property would allow Kazakhstan to increase total tax revenues without creating large economic distortions and/ or to rebalance its tax mix away from more distortive taxes such as the CIT. However, before such a policy could be adopted, Kazakhstan will need to develop a fiscal cadastre that contains information about the characteristics of the property as well as put systems in place that determine the market value of immovable property and make them available for taxation purposes.

Figure 4.24. Revenues from taxes on property are low

Revenues from taxes on property as a share of total tax revenues, Kazakhstan and OECD, 2005 - 2017

4.6.3. Increase scope of carbon prices

The introduction of a carbon tax as a price floor could be complementary to the existing trading system in Kazakhstan. The carbon pricing should be broad based. Countries can choose to price carbon emissions through fuel or carbon taxes or emissions trading systems (ETS). Carbon prices put a price on emissions, so they increase the price of carbon-intensive products and assets relative to low-carbon ones. This steers behaviours and investments towards the latter and reduces emissions. Carbon prices that apply to a broad emissions base provide stronger incentives for investment in clean technology. Kazakhstan currently uses an ETS as a carbon pricing instrument, which it relaunched in 2018 following its suspension in 2016 (after the fall in oil prices). In the absence of a price stability support measure, ETS tend to produce volatile carbon prices that can fail to send a stable signal to invest in low-carbon options (Flues and Van Dender, 2020[26]). ETS have a number of advantages, but given their base is emissions, they generally involve higher administrative and compliance costs than carbon pricing via the more widely used fuel-based approach (for more details, see Box 4.7). To encourage the decarbonisation of the economy (including industrial processes, household energy use, transport or electrical generation) and incentivise investments in low-carbon technologies the emissions base covered by carbon prices should be as broad as possible.

An increasing number of jurisdictions levy explicit carbon taxes in addition to emissions trading systems, for example as carbon price floors or to expand the base to which carbon prices apply (OECD, 2019[27]) (OECD, 2018[28]). While there are different approaches to designing carbon taxes (Box 4.7), if such a tax were introduced in Kazakhstan, it could be integrated with existing excise duties. Indeed, most of the countries that currently have explicit carbon taxes collect them from fuel suppliers in the same way as fuel excise taxes. Countries that follow this fuel-based approach do not tax carbon emissions directly, but put a price on fossil fuels depending on the carbon content of each fuel and convert that price into regular commercial units, for instance by reference to kilogrammes for solid fuels, litres for liquid fuels, and cubic metres for gaseous fuels. This reduces costs linked to measurement of emissions.

Carbon pricing can help to align investment and consumption choices with climate objectives. Broadening the carbon pricing base to sectors that are currently not covered (OECD, 2019[27]) (OECD, 2018[28]) would increase revenue while providing signals to reduce harmful emissions over time. Evidence from several countries suggests that transferring a part of the additional revenues from higher carbon or energy taxes on heating fuels and electricity to poor households is often sufficient to alleviate energy affordability concerns and offset the additional burden these taxes put on low-income households (Flues and van Dender, 2017[29]). Gradual price increases can also alleviate concerns over industry competitiveness, while helping to strengthen the ability of industry to thrive in a context of decarbonisation. Too many energy users do not pay the energy and carbon taxes needed to curb dangerous climate change, even when comparing carbon price signals against a low-end carbon benchmark of EUR 30 per tonne of CO2 . This benchmark is unlikely to reflect the climate damage caused by a tonne of CO2 emitted at present, and will not be sufficient to meet the objectives of the Paris Agreement (OECD, 2019[27]).

Box 4.7. The different approaches to designing carbon taxes

Fuel approach

The most common approach to carbon taxation has been to levy carbon taxes on specific fossil fuels, primarily oil, gas and coal, and their derivative products. Countries that follow this fuel-based approach do not tax carbon emissions directly, but put a price on fossil fuels depending on the carbon content of each fuel and convert that price into regular commercial units, for instance by reference to kilogrammes for solid fuels, litres for liquid fuels, and cubic metres for gaseous fuels.

Under this approach, carbon taxes are collected from fuel suppliers in the same way as pre-existing fuel excise taxes, which lowers administrative and compliance costs. Indeed, the countries that have introduced a carbon tax have generally added it to already existing excise duties, either as part of the general excise duty (e.g. in France) or as a separate tax (e.g. in the Nordic countries).

Direct emission approach

The other approach to carbon taxation consists in taxing carbon emissions directly, by relying on the direct measurement of emissions from certain types of stationary installations/facilities. Given the need to measure (or calculate) emissions, these carbon taxes often apply to emitters above a certain emissions threshold or to installations that fulfil certain technological criteria. Countries that pursue such an emissions-based approach with their carbon taxes include for instance Chile, Estonia and Latvia.

One of the advantages of emissions-based approaches is that they can readily be extended to non-energy and non-carbon emissions, e.g. in agriculture or industry. On the other hand, administrative and compliance costs tend to be somewhat higher than with fuel-based approaches. Whether such differences in administrative and compliance costs are relevant in practice may depend on pre-existing reporting obligations for other purposes. Indeed, the additional effort of reporting carbon emissions for tax purposes may be negligible for facilities that already have reporting obligations for other reasons (e.g. requirements to measure emissions by Integrated Pollution Prevention and Control regulations or by national environmental codes). More generally, administrative and compliance costs become relatively less significant as carbon tax levels increase.

In practice, the choice between fuel-based and emissions-based carbon taxes will also be influenced by political and legal/ constitutional considerations. For instance, in many countries, fuel-based carbon taxes fall under the responsibility of finance ministries, whereas emissions-based carbon taxes (and emissions trading systems) may be under the remit of environment ministries.

Source: OECD Taxing Energy Use 2019 and UN Committee of Experts on International Cooperation in Tax Matters.

References

[8] Alpysbaeva, S. et al. (2020), “Transition to the Progressive Scale of Individual Income Tax in Kazakhstan: Opportunities and Limitations”, Studies on Russian Economic Development, Vol. 31/1, pp. 120-127, http://dx.doi.org/10.1134/S1075700720010025.

[4] Andrews, D., A. Caldera Sánchez and Å. Johansson (2011), “Towards a Better Understanding of the Informal Economy”, OECD Economics Department Working Papers, No. 873, OECD Publishing, Paris, https://dx.doi.org/10.1787/5kgb1mf88x28-en.

[5] Boonzaaier, W. et al. (2017), How do small firms respond to tax schedule discontinuities? Evidence from South African tax registers, https://vatt.fi/documents/2956369/4541479/wp85.pdf/9a45c0bd-7221-4f89-a1d8-f04e89a91c66/wp85.pdf.pdf (accessed on 10 December 2018).

[18] Brys, B. et al. (2016), “Tax Design for Inclusive Economic Growth”, OECD Taxation Working Papers, No. 26, OECD Publishing, Paris, https://dx.doi.org/10.1787/5jlv74ggk0g7-en.

[13] Colin, C. and B. Brys (2019), Fiscal Federalism Working Paper 30: Population ageing and sub-central governments long-term fiscal challenges and tax policy reform.

[26] Flues, F. and K. Van Dender (2020), Carbon pricing design: Effectiveness, efficiency and feasibility: An investment perspective, https://www.oecd-ilibrary.org/taxation/carbon-pricing-design-effectiveness-efficiency-and-feasibility_91ad6a1e-en (accessed on 21 July 2020).

[29] Flues, F. and K. van Dender (2017), “The impact of energy taxes on the affordability of domestic energy”, OECD Taxation Working Papers, No. 30, OECD Publishing, Paris, https://dx.doi.org/10.1787/08705547-en.

[7] Institute of Economic Research (2018), Assessing the macroeconomic effects of reforming individual income tax.

[2] Ivanova, A. et al. (2005), The Russian Flat Tax Reform IMF Working Paper Fiscal Affairs Department The Russian Flat Tax Reform.

[25] Johansson, Å. et al. (2008), Unclassified ECO/WKP(2008)28 JT03248896 Document complet disponible sur OLIS dans son format d’origine Complete document available on OLIS in its original format, http://www.oecd.org/eco/working_papers (accessed on 6 February 2020).

[6] Kleven, J. (2016), “Bunching”, Annual Review of Economics, http://dx.doi.org/10.1146/annurev-economics-080315-015234.

[27] OECD (2019), Taxing Energy Use 2019: Using Taxes for Climate Action, OECD Publishing, Paris, https://dx.doi.org/10.1787/058ca239-en.

[22] OECD (2019), The Role of Digital Platforms in the Collection of VAT/GST on Online Sales, OECD Publishing, Paris, https://dx.doi.org/10.1787/e0e2dd2d-en.

[28] OECD (2018), Effective Carbon Rates 2018: Pricing Carbon Emissions Through Taxes and Emissions Trading, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264305304-en.

[12] OECD (2018), OECD Reviews of Health Systems: Kazakhstan 2018, OECD Reviews of Health Systems, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264289062-en.

[17] OECD (2018), Reshaping the Personal Income Tax in Slovenia.

[9] OECD (2018), Taxation of Household Savings, https://www.oecd-ilibrary.org/docserver/9789264289536-en.pdf?expires=1555420390&id=id&accname=ocid84004878&checksum=56FE29796CE5246BB6EF8F8EB424A435 (accessed on 16 April 2019).

[3] OECD (2017), Building Inclusive Labour Markets in Kazakhstan: A Focus on Youth, Older Workers and People with Disabilities, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264273023-en.

[15] OECD (2017), Health at a Glance 2017 - OECD Indicators - en - OECD, https://www.oecd.org/social/health-at-a-glance-19991312.htm (accessed on 5 August 2020).

[11] OECD (2017), Revenue Statistics - classification of taxes, Interpretative guide and methodology - OCDE, http://www.oecd.org/fr/ctp/politiques-fiscales/revenue-statistics-methodology-guide-and-classification-system.htm (accessed on 5 August 2020).

[24] OECD (2017), Technology Tools to Tackle Tax Evasion and Tax Fraud - OECD, https://www.oecd.org/tax/crime/technology-tools-to-tackle-tax-evasion-and-tax-fraud.htm (accessed on 9 September 2020).

[16] OECD (2015), Fiscal Sustainability of Health Systems: Bridging Health and Finance Perspectives, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264233386-en.

[10] OECD (2004), OECD Tax Policy Studies Recent Tax Policy Trends and Reforms in OECD Countries OECD Tax Policy Studies Recent Tax Policy Trends and Reforms in OECD Countries SourceOECD@oecd.org Recent Tax Policy Trends and Reforms in OECD Countries, http://www.SourceOECD.org, (accessed on 28 February 2020).

[20] OECD/KIPF (2014), The Distributional Effects of Consumption Taxes in OECD Countries, OECD Tax Policy Studies, No. 22, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264224520-en.

[1] Shum, C., J. Fay and G. Man Ching (2017), “Individual income tax : a cross-continental comparison”, World Journal of Social Sciences, https://commons.ln.edu.hk/sw_master/7269 (accessed on 11 February 2020).

[14] State of the Nation (2019), President of Kazakhstan Kassym-Jomart Tokayev’s State of the Nation Address, September 2, 2019, http://www.kisi.kz/index.php/en/n-nazarbayev-address-to-the-nation/3822-2-sentyabrya-2019-g-poslanie-prezidenta-respubliki-kazakhstan-kasym-zhomarta-tokaeva-narodu-kazakhstana-2 (accessed on 3 September 2020).

[23] World Bank (2015), An inquiry into VAT refund and fraud, joint economic research programme.

[19] World Bank (2015), An inquiry into VAT refund and fraud, Joint Economic Research Programme.

[21] World Bank (2008), A strategic plan for increasing the neutrality of the tax system in non-extractive sectors, http://siteresources.worldbank.org/INTKAZAKHSTAN/147271-1090539544329/21824119/KZ_Tax_Report_Vol_1_19Jun08_eng.pdf (accessed on 7 August 2019).

Notes

← 1. Taxes on payroll are generally defined as taxes paid as a proportion of payroll that do not give entitlement to social benefits. Payroll taxes are relatively uncommon in OECD countries.

← 2. The calculation is as follows: [PIT + payroll tax + total SSCs] / [payroll fund + payroll tax]).

← 3. The average inflation rate over the period is based on data from the Asian Development Bank data library.

← 4. The mean wage is taken from the Statistics Committee for the full year 2018.

← 5. Or, as another example if an additional tax bracket was introduced with a lower PIT rate for those in low incomes.

← 6. For example, in the KZT 180 000 – KZT 210 000 income range, a taxpayer is assumed to earn KZT 195 000. Subtracting the obligatory pension payment and basic allowance for 2017 gives taxable income of about KZT 151 000.

← 7. Based on data for 2019 without including the exemption for low-income employees.

← 8. The definition of social security contributions (SSCs) used is that described in the OECD Interpretative Guide.

← 9. They were previously 5% before 2018.

← 10. Data based on World Bank development indicators consumer price inflation for 2017. Uzbekistan not included and Tajikistan rate is for year 2016.

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Countries

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Publications

Publications

Featured publications

Data

Data

Featured data

News & Events

News & Events

Featured Events

About OECD

About

Engage with us

Work with us

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment