Printer Friendly

The long-term effects of capital gains taxes in New Zealand.

1. Introduction

This paper develops a micro-founded macroeconomic model to analyse the long-term effects of capital gains taxes on New Zealand's residential property markets. The model is a version of the overlapping generations lifecycle model pioneered by Modigliani and Brumberg (1980), and adapted to analyse housing issues by Ortalo-Magne and Rady (1998, 2006). The heart of the model is a series of dynamic, forward-looking microeconomic maximization problems in which agents make choices about the type of housing in which they live, how much they consume and save, and how much they borrow and lend. These agents, who differ by income, age, and wealth, have choices over whether to rent or buy, to live in large or small houses, or to share housing with other people. They face realistic bank imposed constraints on the amount they can borrow and the repayment schedule they face if they a purchase a house, and they face a tax system that closely reflects that prevailing in New Zealand. Particular attention is paid to the various ways that taxes on housing income differ according to whether one is an owner-occupier of housing or a landlord. The model is macroeconomic as dynamic steady-state paths for house prices and rents are determined endogenously, and reflect the interaction of decisions by households, landlords, and a construction sector to demand or supply housing. The solution of the model is a set of housing prices and ownership patterns that depend on fundamental parameters such as interest rates, construction sector supply elasticities, the inflation rate, and the particulars of the tax system.

The paper examines how these prices and ownership patterns change as taxes and the inflation rate change, and uses these results to evaluate the consequences of different possible tax systems. Four variants of a capital gains tax regime are examined. While all four are accruals based, they differ according to whether owner-occupied housing is taxed or exempt, and whether capital gains are treated as income or simply taxed at a flat rate. Many of the results of the four variants are similar, although there are important differences, particularly in the amount of revenue that is raised by the tax. In general, when the inflation rate is moderate, capital gains taxes lead to an increase in rents, an increase in the home-ownership rate, a small reduction in number of large houses in the economy, and an increase in the net foreign asset position. However, the effects on economic welfare are ambiguous, for many low-income households suffer a welfare loss from the increase in rents. The simulations suggests the welfare consequences will be worse for low income households if owner-occupied housing is exempt from the tax, although this result is dependent on the revenue from a capital gains tax being refunded to households (including low income households) through a reduction in the GST rate.

The paper is organised as follows. Section 2 outlines the structure of the model. Section 3 discusses the results of the simulations, beginning with an exploration of the welfare consequences of the effects of inflation on the housing market, and concluding with a discussion of the welfare consequences of different capital gains tax systems. Conclusions are offered in Section 4.

2. An intergenerational model of housing demand

2.1. The basic framework

The model is an extension of the model used by Coleman (2008) to analyse the effect of inflation, credit constraints and New Zealand's tax system on the housing market. In turn, it is based on the overlapping generations housing model of Ortalo-Magne and Rady (1998, 2006). The basic structure of the model, outlined below, has four key parts: the demand for rental housing; the demand for owner-occupied housing; the supply of rental housing; and the total supply of housing. The mathematical details of the model are described in the longer working paper version of the paper. (1)

The demand for housing is based on an intertemporal utility maximisation model of consumer demand applied to a large number of agents who differ by age, income, and wealth. In the model, there are four cohorts each containing 400 agents, with each agent passing through four distinct stages (two young stages, one middle-aged stage, and one stage in retirement) before dying. The agents have different exogenously determined labour income, which follows a life-cycle pattern. The agents consume a single non-storable good, pay tax, save for retirement, and have choices over different types of housing at each stage of their lives--whether they share housing with other agents, rent a small house (an apartment), buy a small house or buy a large house. The agents choose their most preferred housing options, given their age, wealth and after-tax incomes, the cost of renting or buying different houses, and their ability to raise a mortgage. Agents can borrow or lend at exogenously determined interest rates, although young agents face bank-imposed credit constraints limiting the amount they can borrow. In the last period of life agents consume all wealth except their house, which is inherited by a younger generation.

The model is dynamic and house prices and rents can change through time. Indeed, when choosing their housing options agents take into account both the rate at which house prices appreciate and the tax treatment on any capital gains that they make. Strictly speaking, in the model house prices and rents comprise two parts: a price level at some base period (t = 0); and a price (or rent) appreciation rate. The model calculates the rate of property price appreciation as part of the process by which it calculates equilibrium prices; while it is normally the general inflation rate, it does not need to be.

Agents are assumed to be forward looking, so when they choose housing in a particular period they take into account not only their current income and current housing prices, but their remaining length of life, future house prices, their future income stream, and their desired future housing patterns. The model includes a careful representation of the conditions imposed by banks on those obtaining mortgage finance to purchase a house, including realistic constraints on the minimum deposit and the maximum mortgage repayment to income ratio. These constraints mean that young households may choose to rent rather than buy a house when inflation and nominal interest rates are high, because they cannot obtain suitable financing.

The utility maximisation model generates housing demand for each of the agents during each of their life stages, for a given set of rent and house price paths. These different housing demand functions are then aggregated together. The resulting aggregate demand functions describe how the demand to rent, the demand for small houses, and the demand for large houses varies as a function of the rent and the price of each type of house, as well as all the basic parameters of the model such as income, interest rates, and tax rates. The basic parameters are listed in Table 4.

Rental accommodation is supplied by agents who become landlords. It is assumed that entry into the rental sector is competitive, so landlords bid for houses and set rents at levels that leave them indifferent between the after-tax returns from lending money and the after-tax returns from investing in residential property. The marginal competitive landlord is assumed to be a middle aged agent who is on the top marginal income tax rate. Particular care has been taken to ensure that taxes in the model replicate the taxes currently imposed on housing in New Zealand. If house prices increase over time, a capital gains tax will lower returns to landlords, and, for a given level of house prices, rents will be higher than they would otherwise have been.

Prices are determined endogenously in the model by equating the total demand for different types of houses with the supply of different types of houses. Cost functions describing the costs of building large and small houses are specified exogenously in the model, and can take any form. In this paper, I focus on the case that there are separate upward sloping supply curves for the quantity of large and small houses, each with approximately unit elasticities. An elasticity of 1 is broadly consistent with the long run increase in prices and the quantity of houses in New Zealand between 1960 and 2005. Two different parameterisations that reflect house prices that are relatively high or relatively low in comparison to income because of high or low construction costs are examined. Several other combinations of supply elasticities have also been analysed, including the cases when the supply of both classes of houses are either perfectly elastic or perfectly inelastic, and the case that the supply of small houses is more elastic than the supply of large houses.

A solution to the model is obtained by finding a set of prices that equate the aggregate demand for different types of housing with the aggregate supply of these types of housing. The prices are solved using a complex numerical routine that calculates the housing demand for each of the 1600 different households for a set of prices, and then chooses a sequence of prices until a set is found at which aggregate demand equals aggregate supply. For this equilibrium set of prices, overall demand patterns are calculated.

As Coleman and Scobie (2009) argue, the effect of taxes, inflation, and interest rates on the housing market depends on a few crucial elasticities, including (i) the elasticity of the total supply of houses to the price of houses (the elasticity of the supply of housing); (ii) the elasticity of the supply of rental housing with respect to rents; (iii) the elasticity of the demand for rental housing with respect to rents and the prices of houses; and (iv) the elasticity of the total demand for housing with respect to rents and the price of houses. The elasticity of the supply of housing with respect to prices is set equal to 1 in the main versions of the model discussed below, but the results have also been analysed when this elasticity is near zero or very large. The supply of rental housing is perfectly elastic with respect to rent, because landlords are assumed to be perfectly competitive and to supply rental housing until the long run after-tax return on rental accommodation is equal to the after-tax return on interest income. The demand elasticities are not directly imposed, but are implicitly derived from the consumer maximisation problem and depend on the basic parameters in the model. These elasticities can have a major effect on the model's results and warrant further discussion.

The elasticity of the demand for rental accommodation to housing rentals and the prices of houses is a measure of the extent to which households are prepared to substitute between renting and home-ownership. This will depend on the relative utility households get from sharing, renting, or owning a house. These parameters are explicitly specified in the model; typically, households are assumed to gain less utility from renting rather than owning, because they can shape an owned house in their own image, and less utility from living in shared accommodation than living by themselves. The substitutability between rental and owned accommodation will be greater the smaller the differences between renting and owning. The more willing households are to substitute between rented and owned housing, the less will be the utility loss from various housing market imperfections or policy interventions.

The elasticity of total housing demand, with respect to house prices or rents, measures the extent to which new households form when prices change as adult children leave home, or as young adults form households by themselves rather than sharing with a group of others. This elasticity is important as it is the only mechanism by which total housing demand can be altered. The more willing are households to share with others, the smaller are the price changes necessary to equilibrate the housing market.

The model analyses the way households climb a 'housing ladder' over the course of their lives. In large part their ascent can be characterised by two factors: the ultimate height they reach and the speed that they attain that height. The ultimate height is largely determined by the ratio of lifetime income to the user-cost of housing: people with higher lifetime incomes will be able to afford larger houses than people with low lifetime income. In the parameterisations studied a majority of people choose a large house in middle age, partly because the tax system favours home ownership as imputed rent is not taxed. The speed of ascent is determined by the steepness of the earnings profile, interest rates and the availability of credit from banks, and the tax incentives facing both households and property investors. In equilibrium, the mix of small and large houses in the economy is determined both by the length of time spent climbing the housing ladder and the peak rung a household attains. Policies that extend the length of time climbing the housing ladder do not, however, necessarily reduce the demand for large houses because an agent can use the money saved from living in a small house while young to live for longer in a large house while old.

2.2. Modelling capital gains taxes

When the inflation rate is positive, income from interest paying assets is taxed more heavily than income from other forms of capital assets because the inflation component of nominal interest earnings is taxed, while capital gains are not. The asymmetry of this treatment means that the tax system generates an incentive for agents to borrow and invest in assets that appreciate over time, potentially inducing agents to invest too heavily in residential housing assets, and lowering home ownership rates among younger and lower income agents. One possible method to eliminate the asymmetrical tax treatment would be to exempt the inflation component of interest income from income tax, as it is not income. Another potential method would be to impose a capital gains tax on residential property and other assets.

Four different capital gains tax schemes are considered. The first scheme treats capital gains as income, and taxes these gains at a taxpayer's marginal income tax rate. Owner-occupied housing is exempt. As landlords are assumed to be higher income, middle aged agents, the applicable tax rate is the top marginal tax rate, 33%. (2) The second scheme also taxes capital gains at the standard marginal income tax rates of 20 or 33%, but in this case owner-occupied housing is taxed. The third scheme imposes a flat rate capital gains tax of 20%, with an exemption for owner-occupied housing. This tax scheme is similar to that which operates in the United States of America. The fourth scheme is similar, except all housing including owner-occupied housing is liable to a flat capital gains tax of 20%.

The model's results when the capital gains of owner-occupied housing are taxed are conceptually problematic. Almost no countries apply capital gains taxes to owner-occupied housing, for a variety of economic and political reasons. One reason is that these taxes are usually only imposed on realised gains when a house is sold, a rule that might deter households from moving from one location to another, perhaps in response to work opportunities. Another reason concerns the financial hardship such a tax could cause when a household dissolves, perhaps because of divorce. These negative effects, which may have first-order welfare consequences, cannot be modelled in this paper and are ignored. To avoid these sorts of issues, it is assumed that if a capital gains tax is introduced without an exemption for owner-occupied housing, it is imposed on an accruals basis: that is, the household is liable for capital gains tax each year, whether or not it is realised through the sale of the house. These results are used to provide a reference case for the effect of a capital gains tax.

The rate of property price appreciation is an outcome of the model, and property prices appreciate in real terms over time when there is population or income growth, unless the supply of housing is perfectly elastic. In the results presented below, however, there is neither income growth nor population growth and so construction costs and property prices increase at the inflation rate. In these circumstances, a capital gains tax only taxes the increase in nominal housing wealth that is due to inflation and thus it reduces the distortion that arises from the asymmetrical taxation of interest income and other assets. Nonetheless, the tax system remains non-neutral with respect to the inflation rate because tax rates on real capital income are an increasing function of the inflation rate when the inflation component of interest income is taxed.

The revenue raised by taxing capital income and/or capital gains depends on both the tax rates and the inflation rate. In the model, any additional revenue raised from changes in the tax system or changes in the inflation rate are refunded through a change in the Goods and Services Tax (GST rate, so that the amount of tax raised is invariant to the tax system.

3. Results

3.1. The effect of inflation

3.1.1. Price and distributional effects of inflation

Inflation has three major effects on the housing market. First, it increases the rate at which property prices increase in nominal terms, generating nominal capital gains for the owners. Since nominal interest earnings rather than real interest earnings are taxed, there is an incentive for owners of capital to invest in residential housing when the inflation rate is high. This incentive applies to both landlords and owner-occupiers, so by itself inflation does not necessarily lead to a decline in the homeownership rate, although it may lead to over-investment in residential housing.

Secondly, inflation may lead to a reduction in nominal rents. This is because (i) after-tax real returns from interest earning assets decline as the inflation rate increases, as the inflation component of interest income is taxed, and (ii) landlords get a portion of their return as capital appreciation, and are prepared to pay more for houses or to accept less rent in order to become landlords. The balance between lower rents and higher prices will depend on the supply elasticity of housing. When supply is relatively elastic, and new construction limits the amount property prices increase, rents will tend to fall.

Thirdly, inflation exacerbates the credit constraints facing agents who borrow to buy houses. This is because bank imposed restrictions on the amount households can borrow are rarely adjusted for inflation, even though nominal interest rates increase when the inflation rate rises. If banks do not increase the amount credit-constrained households can borrow when nominal debt servicing payments increase, it becomes more difficult for these households to purchase houses (Modigliani, 1976; Kearl, 1979). In addition, if rents fall, inflation makes it attractive for young, credit constrained agents to rent rather than purchase and home ownership rates are likely to decline.

Tables 1(a) and 1(b) show how inflation affects long-term housing market outcomes in the model when there is no capital gains tax and there is an elastic supply of housing. In Table 1(a) construction costs and house prices are approximately 25% higher than in Table 1(b). The tables show how prices and rents, the number of houses, the fraction of people owning, and the steady-state level of net financial assets vary with the inflation rate. In both cases, a 2% increase in the inflation rate leads to a 6% reduction in rents, a 0.8% reduction in the number of houses in the economy, a 3-4 percentage point increase in the fraction of the population renting, and a 2-3 percentage increase in the fraction of the housing stock that is owned by landlords and leased. More people rent at all ages. Irrespective of construction costs, there is a reduction in the number of small houses as the combination of falling rents and tighter credit constraints induces more young households to share, reducing aggregate demand for housing. The effect on the number of large houses in the economy differs in the two cases. When construction prices are high, rising inflation increases the total demand for large houses despite falling demand among younger cohorts: there is an increase in the demand by older households, because of the tax advantages of using a house as a saving vehicle. However, when construction costs are low, most people who want to live in a large house can afford to do so for most of their lives, and inflation has a very small effect on the quantity of large houses.

3.1.2. The welfare effects of inflation

Inflation has ambiguous effects on welfare. In keeping with earlier work by Modigliani (1976), Kearl (1979), Feldstein (1996, 1997), and Coleman (2007), agents who wish to purchase a house find inflation tightens credit constraints, because nominal interest rates increase and banks do not change their lending terms and conditions to make an allowance for the way inflation reduces the real value of the nominal outstanding debt. This makes it more difficult for the agents whose real incomes increase over time to smooth consumption, for they have to reduce their consumption to make higher nominal interest payments if they purchase a house. In contrast, those agents who rent benefit from inflation, because it lowers the rent they pay and enables them to spend more while young than they otherwise could.

Whether inflation causes welfare losses or improvements on average depends on the relative size of the populations that rent and own when young. In turn, this depends on the ratio of house prices to incomes. When construction costs are high, a large fraction of young people will wish to share accommodation with others rather than live in a house alone. In this case inflation increases their welfare, because it reduces the negative effects of borrowing constraints that prevent them from smoothing their consumption through time. When construction costs are lower, or social norms make it normal for young adults to either live with family or live by themselves, inflation lowers welfare by making it harder for young agents to buy their first homes. In the parameterisations analysed in this paper, inflation is, on balance, welfare enhancing because there are more agents who benefit from lower rents than there are agents who suffer from higher interest payments at the start of the mortgage. In the real world, which of these two competing effects dominates is an empirical matter. The answer will depend in part on the social mores and conventions of society, particularly the acceptability of sharing housing with non-family members.

Inflation causes one additional welfare effect in the model: it changes the equilibrium number of houses and house prices, which changes the user costs of housing. If inflation leads to an increase in total housing demand, because lower rents entice adult children to leave home earlier, house prices will rise. This tends to lower the welfare of other agents, because the user cost of housing rises and these agents have less to spend on other goods. (3) This effect is an example of the negative pecuniary externality that occurs when agents disregard the effect of their actions on the prices paid by other members of the economy. In contrast, if total housing demand falls in response to inflation, house prices fall, and the welfare of other agents increases. In the parameterisations analysed in this paper, inflation lowers total housing demand, so there is a small positive pecuniary externality that improves the welfare of all agents because of lower house prices.

3.2. The effect of capital gains taxes with an exemption for owner-occupied housing Tables 2(a) and 2(b) show the long-term effects of capital gains taxes when the inflation rate is 2%, owner-occupied housing is exempt, and there is an elastic housing supply. The parameterisations are consistent with Tables 1(a) and 1(b), with construction costs and house prices approximately 25% higher in Table 2(a) than in Table 2(b). The first columns show the equilibrium values of various aspects of the housing market under current tax regulations. The second and third columns shows how these values change when capital gains are treated as income and taxed at marginal income tax rates (20%, 33%) and when capital gains are taxes at a flat rate 20% rate. Note that most landlords are high income agents, so the former scheme is effectively a flat rate capital gains tax with a 33% tax rate. The fourth column shows how these values change when the inflation component of interest income is exempt from income tax. In each case, there is neither income nor population growth and property prices appreciate at the inflation rate. Consequently, when the inflation rate is zero a capital gains tax has no effect.

3.2.1. Taxing capital gains on leased property at marginal income tax rates

The main effect of the (20%, 33%) capital gains tax is to increase rents by approximately $1300 or 11% (Table 2(a)). (4) There is also a small increase in house prices, by 0.6-0.8%. The latter occurs because the demand for property increases: the increase in rents makes renting less attractive and there are a number of agents who cease living in shared rental accommodation and purchase and live in a house by themselves.

The decline in renting is most noticeable among older households. Among younger (cohort 0 and 1) households, renting only declines modestly, for credit constraints are sufficiently tight on most low income agents that renting is still more attractive than home ownership, particularly as most of these agents share rental housing and thus only experience half of the rent increase. In both Tables 2(a) and 2(b), 2.5-3.5% of cohort 0 and cohort 1 cease renting. In contrast, middle-aged and retired households almost completely cease renting, because the capital gains tax raises the long-run cost of renting above the cost of owning as only landlords pay the capital gains tax. The total effect of the capital gains tax on the rental market depends on the number of people initially renting, which depends on construction costs and house prices. When construction costs are high, a large number of agents rent under the current tax rules and a capital gains tax reduces the fraction of agents renting by over six percentage points; when they are low, fewer middle aged and older agents rent, and the capital gains tax only reduces the fraction renting by 3%.

There are two other economic effects. First, the simulations indicate the capital gains tax raises little revenue, for the GST rate only declines by 0.1 percentage points. This is partly because the capital gains tax reduces the number of rental houses, and thus leads to a reduction in the income tax paid by landlords on their rental income. Secondly, the simulations suggest that there is a small increase in the net financial asset position of the economy. The amount is larger when construction costs are high rather than low, and reflects the increase in saving that occurs as some households switch from being lifetime renters to middle-aged home-owners due to the increase in rents.

3.2.2. Taxing capital gains on leased property at a constant 20% rate

The results are qualitatively similar if the capital gains tax is applied at a flat 20% rate, except the effects are quantitatively smaller as the average tax rate is 20% rather than 33%. The increase in rents is only 60% as large, and there is a correspondingly smaller decrease in the fraction of agents that rent and the fraction of houses that are leased. Interestingly, more revenue is raised under the lower flat rate capital gains tax when construction costs are high, as there is a much smaller decline in renting. Nonetheless, the amount of tax raised under either capital gains tax regime is small and the reduction in GST is less than 0.2% in either case.

3.2.3. Taxing capital gains on leased property: results for other supply functions

When the supply of small house is elastic but the supply of large houses is inelastic, the results are similar to the case when both the supply curves are elastic. Once again, the primary effect of the capital gains tax is to increase rents, reduce the fraction of the population that is renting and the fraction of the housing stock that is leased, and increase the total housing stock.

When both supply curves are inelastic, the results are a little different. In Table 2(c), the number of flats and houses is constant and, because there is an overall shortage, prices are high. As before, a capital gains tax raises rents and reduces the number of people renting, particularly amongst those who are middle aged or retired. However, house prices increase quite sharply, by 4 or 5%. The house price increase is needed to reduce the total demand for housing, because an increase in rents without an increase in house prices leads to a reduction in the number of agents sharing rental accommodation and an increase in the total demand for housing. The only way to reduce the total demand for housing is to raise house prices, and make it attractive to share. When the supply of housing is elastic, the price rise is not necessary, as new houses are built to meet the additional demand. If a rise in rents leads to a reduction in total demand, because some young agents respond to the increase in rents by moving back to their parents' home, a capital gains tax could lead to a fall in house prices as well as an increase in rents. Consequently, the way prices would behave in New Zealand if a capital gains tax were introduced will depend on the size of the elasticity of total demand for housing to rent.

3.3. Exempting the inflation component of interest from income tax

The fourth columns of Tables 2(a)-2(c) show what happens if neither capital gains nor the inflation component of interest income were taxed. In all of the housing supply versions considered, the effects on rents, prices, and home ownership rates are similar to what happens if a flat 20% capital gains tax regime with an exemption for owner-occupiers were introduced. There are two differences, however. Because tax revenue declines slightly when the inflation component of interest income is exempted from income tax, the GST rate has to be increased slightly, rather than cut. The increase is always less than 0.2 percentage points, however, partly because the loss of tax on interest income is offset by a reduction in the deductions allowable against rental income. Secondly, there is a larger increase in the net financial asset position than in either of the capital gains tax regimes considered. This is because exempting the inflation component of interest income from tax raises after-tax real interest rates, encouraging saving and capital accumulation among working age agents: Since a capital gains tax does not affect after tax interest rates, after-tax returns are higher when the inflation component of interest is tax exempt than when a capital gains tax is introduced.

3.4. The effect of an accrual capital gains regime applied to all agents

Table 3 shows what happens to the housing market if capital gains taxes are applied to all households on an accruals basis, either as a flat rate (20%) or at marginal income tax rates (20%-33%).The results are for the case that the housing supply is elastic and construction costs are high, but the results for other housing supply parameters are qualitatively similar.

The capital gains tax leads to an increase in rents. When the capital gains tax rate is a flat 20%, there is little effect on the quantity of housing rented, however, because owner-occupiers are also liable for capital gains tax, so the cost of owning a house rises by a similar amount. There is a significant switch from large houses to small houses, however, as the capital gains tax raises the user cost of large houses by more than the user cost of small houses. This reduces the demand for large houses at all ages. The substitution between large and small houses also occurs because of a sizeable drop in the GST rate that makes the consumption of goods relatively more attractive than the consumption of housing. The capital gains tax revenue is much larger than when owner-occupied housing is exempt, and the GST rate declines by over two percentage points rather than 0.2 percentage points.

When capital gains are taxed as income at marginal income tax rates, there is a substantial decline in the number of middle-aged and retired households renting. This is because rents increase by more than the amount of capital gains tax that would be paid by low income renters, so it is cheaper in the long term for low-income households to purchase rather than rent. Indeed, the reduction in the number of households renting is similar to when owner-occupied housing is exempt from capital gains tax.

In this model, the effects of a capital gains tax applied to all households are very similar to the effects of a flat rate property tax. This is because property prices increase at the inflation rate, so a flat rate capital gains tax end up taxing houses at a rate that is proportional to value. Coleman and Grimes (2009) discuss the effects of introducing a property tax at greater length.

3.5. The welfare implications of capital gains taxes

The above analysis suggests that capital gains taxes raise rents, increase homeownership rates, cause a substitution towards smaller houses, and improve the net foreign asset position because they reduce the distortions caused by the interaction of inflation with the tax system. Whether capital gains taxes raise welfare, however, will depend on two things: the extent to which inflation reduces welfare, because of its negative effects on credit-constrained owner-occupiers: and/or the extent to which inflation enhances welfare because of its positive effects on credit-constrained renters. In the parameterisations of the model studied in this paper, the welfare losses to the renters exceed the benefits to the owners, for there are more young renters than young owners. In this case, a capital gains tax will tend to have negative welfare effects as it raises rents, although this need not be the case.

Figure 1 shows how different tax schemes affect lifetime welfare for people with different income levels when the supply of housing is elastic and construction costs are low. The figure shows the average change in utility for each income decile. (6) Three points stand out. First, a capital gains tax scheme that exempts owner-occupied housing has lower welfare for most people than one that does not. Secondly, a capital gains tax scheme that exempts owner-occupied housing reduces welfare for most low-income people, because of the increase in rents. Thirdly, a capital gains tax scheme that exempts owner-occupied housing has similar welfare properties as a tax scheme that exempts the inflation component of interest income from tax. These three results occurred in most of the parameterisations studied, even though the exact nature of the welfare changes depends on a number of factors such as the housing supply elasticities and the way people inherit property.

Several features of these results are of interest. First, a capital gains tax scheme without an exemption for owner-occupied housing has very similar properties to the flat rate property tax scheme analysed by Coleman and Grimes (2009). As discussed above, this is not surprising, for in the model taxing capital gains on an accrual basis when the inflation rate is constant is like having a flat rate property tax. Since the welfare effects are similar, and since the effects on rents, prices, and home-ownership rates are similar, a flat rate property tax could be a substitute for a capital gains tax if it were believed the inflation rate would continue to be low and stable. Given the political difficulties of introducing a capital gains tax on owner-occupied residential housing in other countries, a flat rate property tax may be an attractive option.

[FIGURE 1 OMITTED]

Secondly, the welfare properties of capital gains regimes that do or do not exempt owner-occupied housing are significantly different, even though they have a similar effect on rents. The differences are caused by two factors. First, when owner-occupied housing is taxed, much more tax is collected. The consequent cut in the GST rate partially compensates renters for the rise in rents, and leads directly to an improvement in their welfare. In addition, there is a reduction in the total demand for property, so house prices fall relative to the case that only landlords pay the tax. This leads to a reduction in the direct user cost of housing to all agents. This provides a gain to all agents in the economy except the first generation, who suffer a capital loss.

Thirdly, a tax regime that exempts the inflation component of interest income from tax has similar welfare properties as a tax regime that taxes capital gains tax on leased residential property. Again, it may be politically easier to introduce such a tax regime than a capital gains tax.

Figure 2 shows the welfare effects of different tax regimes when the supply of housing is elastic but construction costs are high. The results are similar, although there is a downwards spike in the sixth decile that reflects the effect of inheritance arrangements. (7) Once again, the welfare consequences of a capital gains tax that includes owner-occupiers are better than a tax that does not; the welfare consequences of capital gains tax regimes that exclude owner-occupiers are negative for low-income agents because the taxes increase rents; and the welfare consequences of capital gains taxes that exempt owner-occupiers are similar to the welfare consequences of tax regimes that exempt the interest component of interest income from tax. The welfare losses for low income agents are higher when construction costs are high, partly because more people rent but also because rents are higher and thus increases in rents cause more severe cuts in consumption.

[FIGURE 2 OMITTED]

4. Discussion and conclusions

This paper has explored some of the consequences of introducing a capital gains tax on residential property in New Zealand. It has done this in the context of a stylised model that attempts to understand the factors that determine housing market outcomes in the long term. The model focuses on three main factors: the cost of supplying new housing; the financial incentives facing landlords; and the tax and financial incentives facing households as they choose different housing options over the course of their lives.

The model suggests that a capital gains tax will have the following effects: it will lead to an increase in rents; it will lead to a reduction in the number of people renting, and an increase in homeownership rates; it will lead to an increase in the net foreign asset position; and it will lead to a decline in the fraction of large houses in the economy. It is possible that homeownership rates could rise by several percent if a capital gains tax were introduced, with a similar sized increase in the net foreign asset position.

Two other results seem general. First, a capital gains tax that exempted owner-occupied housing would raise little revenue, whereas one that applied to all households would raise enough to allow a sizeable reduction in the GST rate. For this reason, low income households that rent are better off when a capital gains tax does not have exemptions. Secondly, the increase in rents and the increase in home-ownership rates will be larger if the capital gains tax rates on owner-occupied residential property are lower than those on leased residential property, either because the former is specifically exempted from capital gains tax or because landlords typically have higher marginal tax rates than households who typically rent.

Beyond these general outcomes, the paper demonstrates that the welfare implications of a capital gains tax depend a lot on the detailed structure of the economy. It matters whether (i) the supply of housing is elastic or inelastic; (ii) construction costs are high or low; (iii) people prefer to own rather than rent; or (iv) young people respond to rent increases by sharing with more people, or by deciding to buy a house themselves. Indeed, some of these factors matter so much that they determine whether a capital gains tax is largely beneficial or harmful.

It is both a weakness and a strength of the modelling approach that it cannot be more definitive about the welfare effects of a capital gains tax. From a technical perspective, the weakness is clear: a model that delivers different answers when the housing choice set is structured differently makes it difficult to know whether the model's outcomes are robust or contrived. The strength is more subtle: the modelling approach suggests that the welfare effect of different policies depends a lot on several deep parameters in the models, suggesting empirical research on the nature of these parameters is important before policies are introduced.

This paper has ignored many of the practical and political issues that would have to be solved if capital gains taxes were to be introduced. While the simulations of the model suggest a capital gains tax that includes owner-occupied housing has better welfare properties than a capital gains tax that does not, the political and practical difficulties of introducing an accruals based capital gains tax should not be underestimated. Applying a capital gains tax only to realised gains has its own problems, notably the incentives it generates to remain in unsuitable houses or living arrangements in order to avoid the tax. Yet the simulations also suggest that a flat rate property tax has many of the same properties as an accrual based capital gains tax with no exemptions, and if a capital gains tax is desired but not considered practical this may be a suitable alternative. The similarity between these two taxes will be greater if nominal property price appreciation is dominated by inflation rather than real factors, and if the inflation rate is relatively stable.

Most OECD countries that have capital gains taxes exempt owner-occupied housing from the tax and only tax leased residential property when a sale is realised. This is a much more straightforward tax to implement than an accruals based tax, but still removes some of the housing market distortions that arise from taxing differently the inflation component of interest earnings and the inflation component of capital gains. Nonetheless, the simulations suggest the effects of this type of capital gains tax could be largely replicated by exempting the inflation component of interest income from tax, a strategy that may be easier to implement in practice. Such a strategy would have the added advantage that real after-tax interest rates and returns to capital are unaffected by the inflation rate.

The key issue underlying the whole paper is whether the effects of moderate inflation on the housing market largely lowers or improves welfare. In line with earlier work, this paper identifies two ways that inflation affects welfare. First, inflation makes it more difficult for people to purchase a house, or upgrade to a bigger house, because nominal interest rates increase and banks do not change their lending criteria to recognise the way inflation erodes the real value of the existing debt. This is the familiar issue of mortgage tilt, which lowers welfare (Modigliani, 1976.) Secondly, inflation leads to lower rents if interest earning assets are taxed more heavily than capital gains. This improves the welfare of those who are credit constrained and rent. Whether the effect of inflation on the housing market improves or lowers welfare therefore depends on the fractions of the population who find it eases rather tightens the credit constraints they face. In this paper, I have focussed on parameterisations in which inflation raises welfare for many agents, while in earlier work I focused on the case that inflation lowered welfare. In practice, this is an empirical question, the answer of which will depend on the cost of housing and the way agents value renting and home-ownership. If, under the current tax system, inflation lowers rents and raises the welfare of many people, policies that counteract the effects of inflation on the housing market will tend to lower welfare. Conversely, if inflation mainly causes hardship among those who wish to borrow to purchase a house, a capital gains tax improves welfare.

DOI: 10.1080/00779954.2010.492575

Acknowledgements

The author would like to thank Ananish Chaudhuri, Arthur Grimes and seminar participants at the New Zealand Treasury for comments on this paper.

References

Coleman, A. (2007). Credit constraints and housing markets in New Zealand. Reserve Bank of New Zealand Discussion Paper 2007/11.

Coleman, A. (2008). The big costs of small inflation. Motu Economic and Public Policy Working Paper 2008/13.

Coleman, A. (2009). The long term effects of capital gains taxes in New Zealand. Motu Economic and Public Policy Working Paper 09-13. http://www.motu.org.nz/publications/ working-papers/2009

Coleman, A., & Grimes, A. (2009). Fiscal, distributional and efficiency impacts of land and property taxes. Motu Working Paper 09-12.

Coleman, A., & Scobie, G. (2009). A simple model of housing rental and ownership with policy simulations. New Zealand Treasury, mimeo.

Feldstein, M. (1996). The costs and benefits of going from low inflation to price stability. NBER Working Paper 5469.

Feldstein, M. (1997). Capital income taxes and the benefits of price stability. NBER Working Paper 6200.

Kearl, J.R. (1979). Inflation, mortgages and housing. Journal of Political Economy, 87(5), 1115-1138.

Modigliani, F. (1976). Some economic implications of the indexing of financial assets with special references to mortgages. In M. Monti (Ed.), The new inflation and monetary policy (pp. 90-116). London and Basingstoke: Macmillan.

Modigliani, F., & Brumberg, R. (1980). Utility analysis and aggregate consumption functions: an attempt at integration. In Andrew Abel (Ed.), The collected papers of Franco

Modigliani: Volume 2, The life cycle hypothesis of saving (pp. 128-197). Cambridge, MA: The MIT Press.

Ortalo-Magne, F., & Rady, S. (1998). Housing fluctuations in a life-cycle economy with credit constraints. Research paper 1501, Graduate School of Business, Stanford.

Ortalo-Magne, F., & Rady, S. (2006). Housing market dynamics: on the contribution of income shocks and credit constraints. Review of Economic Studies, 73(2), 459-485.

Andrew Coleman *

Motu Economic and Public Policy Research, Wellington, New Zealand

* Email: andrew.coleman@motu.org.nz

Notes

(1.) See Coleman (2009).

(2.) The top marginal tax rate in New Zealand is currently 38%, although it was 33% before 2000. However, most landlords could choose to put a leased property in a trust which is only taxed at 33%. I have chosen to solve the model for a top marginal tax rate of 33% in part because this rate is often seen as a goal by political parties, and in part because of the way landlords can use trusts.

(3.) Annual consumption falls by approximately the real interest rate multiplied by the additional housing cost. Higher house prices also lower the net foreign asset position.

(4.) Small house prices are $225,000, so the capital gain is $4500 when the inflation rate is 2%: if the landlord has to pay 33% of this sum in tax, the rent has to be raised to make the same after-tax return as investing in interest earning assets. The $1300 increase in rent is not exactly equal to 0.33 x $4500 for two reasons. First, in the model, the timing convention is that the landlord is paid rent and pays income tax at the start of the period, but pays capital gains tax at the end of the period. The after-income-tax value of the $1300 rent increase is invested for the length of the period (in this case 12.5 years): in this case the extra interest is approximately the same value as the income tax paid. Secondly, property prices increase by approximately 1% once the CGT is introduced, leading to a 1% or $100 increase in rents.

(5.) While the model has a steady state saving rate of zero, as people run down the assets they accumulate while working when they are retired, the economy's net asset position increases when the saving rate among working age people increases.

(6.) The effects on the welfare of the lowest decile are not shown as they largely reflect the inheritance arrangements in the economy. In these simulations odd-numbered agents receive no inheritance, but even-numbered agents inherit the houses of the two retired agents with the same rank in the income distribution. When there are no capital gains taxes, most decile one agents rent throughout their lives and neither leave an inheritance nor receive one. When a capital gains tax is introduced, many of these agents buy a house, and bequeath it in old age. The logic of the model means that even-numbered agents also inherit one or more houses, and are much better off. Although this effect dominates the welfare calculations for the lowest income decile, this result is not emphasised in this paper.

(7.) In equilibrium many people inherit as well leave a bequest. In this case the changes partly reflect the welfare effects of inheriting a small house rather than a large house.
Table 1(a). The effect of inflation on housing outcomes.
Elastic supply, high construction costs

                         [PI] = 0   [PI] = 1   [PI] = 2   [PI] = 3

Rent                        11900      11650      11250      10850
[P.sup.F],(0) (small)      225200     224500     223600     222700
[P.sup.H] (large)          382900     382400     382000     381400
[N.sup.TOT] (all)/popn      93.9%      93.6%      93.1%      92.8%
[N.sup.F] (small)/popn      53.0%      52.4%      51.4%      50.6%
[N.sup.H] (large)/popn      40.9%      41.1%      41.8%      42.1%
% houses rented             10.7%     l2.2%       13.6%      15.1%
% agents renting            16.1%      17.9%      19.5%      21.2%
% cohort 0 renting          38.0%      41.0%      43.0%      44.0%
% cohort 1 renting          11.0%      13.0%      16.0%      17.0%
% others renting             8.0%       9.0%      10.0%      12.0%
% cohort 1 large            60.0%      58.5%      57.5%      56.0%
GST rate                    12.3%      12.1%      12.0%      12.0%
Net financial
  assets/GDP                  28%        29%        31%        33%

The table shows how rents, house prices, house numbers
(number of houses divided by the population) and the fraction
of the population that rents vary with the inflation rate.
Net financial assets/GDP is total lending minus total
borrowing divided by labour income.

Table 1(b). The effect of inflation on housing outcomes.

                           Elastic supply, low construction costs

Rent                         9650     9350     9050     8700

[P.sub.F](0) (small)       182400   181300   180500   179500
[P.sub.H](0) (large)       325600   324400   323800   322300
[N.sub.TOT] (all)           96.8%    96.4%    96.1%    95.6%
[N.sub.F] (small)           42.8%    42.5%    42.0%    42.1%
[N.sub.H] (large)           54.1%    53.9%    54.1%    53.5%
% houses rented              3.6%     5.3%     6.7%     7.9%
% agents renting             6.6%     8.7%    10.4%    11.9%
% cohort 0 renting          25.0%    26.0%    28.0%    30.0%
% cohort 1 renting           1.0%     4.0%     6.0%     8.0%
% others renting             1.0%     2.0%     3.0%     5.0%
% cohort 1 large            71.5%    71.0%    71.0%    70.0%
GST rate                    12.0%    11.8%    11.5%    11.5%
Net financial assets/GDP      48%      48%      50%      53%

The table shows how rents, house prices, house numbers
(number of houses divided by the population)
and the fraction of the population that rents vary with
the inflation rate. Net financial assets/GDP is total
lending minus total borrowing divided by labour income.

Table 2(a). The effects of capital gains taxes on residential
property; owner-occupied housing exempt, inflation rate = 2%.

                               Elastic supply, high construction costs

                                       Change from introducing ...

                                                 CGT at    Inflation
                                                marginal    part of
                                    Flat rate    rates     interest
                           No CGT   CGT, 20%    20%, 33%   tax exempt

Rent                        11250       +$750     +$1300        +$800
[P.sub.F] (0) (small)      223600      +$1000     +$1800        +$900
[P.sub.H] (O) (large)      382000       +$900     +$1800           +0
[N.sub.TOT] (all)/popn      93.1%        0.4%       0.8%         0.4%
[N.sub.F] (small)/popn      51.4%        0.6%       0.9%         0.7%
[N.sub.H] (large)/popn      41.8%       -0.2%      -0.1%        -0.3%
houses rented               13.6%       -3.8%      -6.2%        -3.5%
agents renting              19.5%       -3.9%      -6.5%        -3.6%
cohort 0 renting            42.5%       -2.0%      -3.5%        -2.3%
cohort 1 renting            15.5%       -1.0%      -2.5%        -1.8%
others renting              10.0%       -6.3%     -10.0%        -5.3%
GST rate                    12.0%       -0.2%      -0.1%          0.1
Net financial assets/GDP    30.6%        2.2%       4.1%         7.0%

The table shows how rents, house prices, house numbers (number
of houses divided by the population) and the fraction of the
population that rents would change if a capital gains tax
exempting owner occupiers were introduced. The inflation rate
is assumed to be 2%. Net financial assets/GDP is total
lending minus total borrowing divided by labour income.

Table 2(b). The effects of capital gains taxes on residential
property; owner-occupied housing exempt, inflation rate = 2%.

                               Elastic supply, low construction costs

                                   Change from introducing ...

                                                 CGT at    Inflation
                                                marginal     part of
                                    Flat rate     rates     interest
                           No CGT   CGT, 20%    20%, 33%   tax exempt

Rent                         9050      +$600     +$1050        +$650
[P.sub.F] (0) (small)      180500     +$1100     +$1500       +$1200
[P.sub.H] (0) (large)      323800     +$1100     +$1600       +$1300
[N.sub.TOT] (all)/popn      96.1%        0.4%                    0.5%
[N.sub.F] (small)/popn      42.0%        0.4%       0.5%         0.4%
[N.sub.H] (large)/popn      54.1%        0.1%       0.1%         0.1
% houses rented              6.7%       -2.3%      -2.6%        -2.6%
% agents renting            10.4%       -2.7%      -3.1%        -2.9%
%, cohort 0 renting         28.3%       -1.8%      -3.0%        -2.3%
% cohort 1 renting           6.3%       -2.0%      -2.5%        -2.5%
% others renting             3.5%       -3.5%      -3.5%        -3.5%
GST rate                    11.5%        0.0%       0.0%         0.3%
Net financial assets/GDP    50.0%        1.2%       1.3%         7.1%

The table shows how rents, house prices, house numbers
(number of houses divided by the population) and the
fraction of the population that rents would change if a
capital gains tax exempting owner occupiers were introduced.
The inflation rate is assumed to be 2%. Net financial
assets/GDP is total lending minus total borrowing divided by
labour income.

Table 2(c). The effects of capital gains taxes on residential
property; owner-occupied housing exempt, inflation rate = 2%.

                                Inelastic supply, high prices

                                  Change from introducing ...

                                   Flat      CGT at     Inflation
                                   rate     marginal      part of
                                   CGT,    rates 20%,    interest
                         No CGT      20%          33%   tax exempt

Rent                      11350   +$1100       +$1850       +$1100
[P.sub.F](0) (small)     225900   +$7000      +$11000       +$6300
[P.sub.H](0) (large)     378300   +$5000      +$16300       +$6200
[N.sub.TOT] (all)/popn    93.2%     0.0%         0.0%         0.0%
[N.sub.F] (small)/popn    50.1%     0.0%         0.0%         0.0%
[N.sub.H] (large)/popn    43.1%     0.0%         0.0%         0.0%
% houses rented           13.3%    -2.6%        -5.1%        -2.4%
% agents renting          19.2%    -2.4%        -4.7%        -2.3%
% cohort 0 renting        42.0%    -0.5%         0.3%        -0.3%
% cohort 1 renting        15.3%     0.8%         0.5%        -0.3%
% others renting           9.8%    -5.0%        -9.8%        -4.3%
GST rate                  11.9%    -0.2%        -0.2%         0.1%
Net financial
  assets/GDP              32.0%     2.1%         3.4%         7.5%

The table shows how rents, house prices, house numbers (number
of houses divided by the population) and the fraction of the
population that rents would change if a capital gains tax
exempting owner occupiers were introduced. The inflation
rate is assumed to be 2%. Net financial assets/GDP is total
lending minus total borrowing divided by labour income.

Table 3. The effects of capital gains taxes
on all households; [pi] = 2.

                                Elastic supply, high prices

                            Change from introducing taxes on ...

                                    All households

                                            Flat    CGT at
                                 No         rate     (20%,
                                CGT     CGT, 20%      33%)

Rent                           11250        +$700    +$1200
[P.sup.F](0) (small)          223600        -$100     +$400
[P.sup.H](0) (large)          382000      -$2,100   -$1,500
[N.sub.TOT]
  (all)/popn                   93.1%         0.0%      0.2%
[N.sub.F]
  (small)/popn                 51.4%         2.5%      2.6%
[N.sub.H]
  (large)/popn                 41.8%        -2.5%     -2.4%
% houses rented                13.6%        -0.4%     -5.4%
% agents renting               19.5%        -0.3%     -5.1%
% cohort 0 renting             42.5%         0.8%     -0.8%
% cohort 1 renting             15.5%        -1.0%     -0.8%
% others renting               10.0%        -0.5%     -9.5%
GST rate                       12.0%        -2.2%     -2.5%
Net financial
  assets/GDP                   30.6%         0.8%      6.5%

                            Elastic supply, high prices

                            Change from introducing taxes on ...

                                landlords only

                               Flat
                               rate       CGT at
                            CGT, 20%   (20%, 33%)

Rent                          +$750       +$1300
[P.sup.F](0) (small)         +$l000       +$1800
[P.sup.H](0) (large)          +$900       +$1800
[N.sub.TOT]
  (all)/popn                    0.4%         0.8%
[N.sub.F]
  (small)/popn                  0.6%         0.9%
[N.sub.H]
  (large)/popn                 -0.2%        -0.1%
% houses rented                -3.8%        -6.2%
% agents renting               -3.9%        -6.5%
% cohort 0 renting             -2.0%        -3.5%
% cohort 1 renting             -1.0%        -2.5%
% others renting               -6.3%       -10.0%
GST rate                       -0.2%        -0.1%
Net financial
  assets/GDP                    2.2%         4.1%

The table shows how rents, house prices, house numbers
(number of houses divided by the population)
and the fraction of the population that rents would
change if a capital gains tax on all houses were
introduced. The inflation rate is assumed to be 2%.
Net financial assets, GDP is total lending minus total
borrowing divided by labour income.

Table 4. Parameterisation of the model.

Parameter                  Description         Value

T                          Length of period    12.5 years

[Y.sup.0.sub.1]            Average income      50000
                             of 25-35 cohort

[[omega].sub.j]            Income              Uniform on
                             distribution      [25000,85000]
[g.sub.i]                  Lifecycle income    {1, 1.5, 1.5,
                             Pattern           0.15 + 20000}

B                          Discount factor     0.97 annualised

{[v.sup.1/2R, [v.sup.R],   Utility from        {0.18, 0.32,
[v.sup.F], [v.sup.H]}        housing            0.35, 0.45}

[K.sub.i]                  Inheritance         {0,0,1,0}
                             timing

[gamma]                    Annual house        0.01
                            maintenance

H                          Mortgage term       25 years

[delta]                    Maximum debt        30%
                             service-income
                             ratio

[THETA]                    Maximum loan to     90%
                             value ratio

[[tau].sup.g *]            Target GST rate     0.14

[[tau].sub.1],             Income tax rates    20%, 33%
[[tau].sub.2],               and threshold     $50,000
[tau] *

Housing supply parameters

[[alpha].sup.F.sub.0]      Elastic             0, 150
[[alpha].sup.F.sub.1]
[[alpha].sup.H.sub.0],     High price          125000,50
[[alpha].sup.H.sub.1],
                           Elastic             -50000, 150
                           Low Price           100000, 50
                           Inclastic           -149m, 100000
                           High Price          -68m, 100000
                           Inelastic           -153m, 100000
                           Low Price           -83m, 100000
                           Flats elastic       0, 150
                           Houses inelastic    -80m, 100000
                           Perfectly elastic   [P.sup.F] = 225000
                                             [P.sup.H] = 370000

Parameter                  Source/Rationale

T                          To approximate work
                             history from age
                             25-75

[Y.sup.0.sub.1]            NZ Census 2001:
                             average male and
                             female earnings,
                             25-35 year olds, are
                             $32800 and $23300
                             respectively

[[omega].sub.j]

[g.sub.i]                  NZ Census, 1966-2001.
                             Based on real
                             lifecycle earnings
                             of cohort turning
                             20 in 1946, 1961.

B                          Arbitrary

{[v.sup.1/2R, [v.sup.R],   Arbitrary
[v.sup.F], [v.sup.H]}

[K.sub.i]                  Arbitrary

[gamma]                    Arbitrary

H                          Standard mortgage term
                             in 1990s

[delta]                    Reflects NZ banking
                             conditions

[THETA]                    Reflects NZ banking
                             conditions

[[tau].sup.g *]            Tax take equals 14% of
                             labour income;
                             arbitrary, but close
                             to NZ rate.

[[tau].sub.1],             Reflects NZ rates
[[tau].sub.2],               in 2000
[tau] *

Housing supply parameters

[[alpha].sup.F.sub.0]      Arbitrary, generates
[[alpha].sup.F.sub.1]        approximately 1%
[[alpha].sup.H.sub.0],       price elasticity for
[[alpha].sup.H.sub.1],       flats.
                           Generates [N.sup.F] = 689
                           [N.sup.H] = 802
                           Generates [N.sup.F] = 665
                           [N.sup.H] = 873
                           [N.sup.H] = 800
COPYRIGHT 2010 New Zealand Association of Economists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:SYMPOSIUM ON PROPOSED CHANGES TO THE NEW ZEALAND TAX SYSTEM
Author:Coleman, Andrew
Publication:New Zealand Economic Papers
Article Type:Report
Geographic Code:8NEWZ
Date:Aug 1, 2010
Words:10048
Previous Article:Population ageing and taxation in New Zealand.
Next Article:Fiscal, distributional and efficiency impacts of land and property taxes.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters