Is capital back? The role of land ownership and savings behavior

Abstract

Wealth inequality is a major political concern in most OECD countries. Under this premise, we analyze different policy instruments in terms of their impact on wealth inequality and output. In a general equilibrium model, we disaggregate wealth in its capital and land components, and savings in their life-cycle and bequest components. Households are heterogeneous in their taste for leaving bequests. We show that governments have considerable freedom in reducing wealth inequality without sacrificing output: Land rent taxes enhance output due to a portfolio effect and reduce wealth inequality slightly. Bequest taxes have the highest potential to reduce inequality, and their effect on output is moderate. By contrast, we confirm the standard result that capital taxes reduce output strongly and show that they only have moderate redistributive effects. Furthermore, we find that using the tax proceeds for transfers to the young generations enhances output the most and further reduces wealth inequality.

Appendices

Model parameters and calibration

See Tables 4 and 5.

Table 4 Benchmark parameters that reproduce observed data on the wealth distribution in OECD countries

Table 5 Comparison of average OECD data and model output

Additional figures and data

See Figures 5, 6, 7 and 8 and Table 6.

Fig. 5

Aggregate composition of assets (cf. Sect. 3.1) under variation of fiscal policy. Fiscal policy that stimulates (hampers) investment in capital will unambiguously increase (decrease) output. While a bequest tax only indirectly affects asset prices, taxes on capital income and land rents have a relatively strong impact on asset prices. As the relative prices of assets change, households react by changing the composition of their portfolio—a portfolio effect takes place. The tax on land rents shifts investment toward capital. The capital stock increases, but the other two inputs, labor and land, do not change, as we assume that they are in fixed supply. Hence, the land rent tax increases output. Raising a capital income tax has the opposite effect. It discourages investments in capital and, thus, leads to output reductions

Fig. 6

The revenue raising potential of fiscal policy depends on the recycling scheme used. For the bequest tax, public revenues are higher the higher the share of transfers to the old. The capital income and the land rent tax are much less sensitive to the transfer scheme used. Note that if all tax revenues are transferred to the old generation ($\delta =1$), the capital income tax is only feasible up to a rate of $70\%$. Figure 4 shows how the choice of the transfer scheme affects output

Fig. 7

Impact of different recycling schemes on output and on the wealth distribution (cf. Fig. 4). The filled points mark the option space for the case in which public revenues are not redistributed (the capital income tax is marked by boxes, the bequest tax by triangles). Thus, transferring all tax revenues to the old generation, for example, would result in a lower steady-state level of output and higher wealth inequality, than if the tax revenues were not recycled to the modeled economy at all

Fig. 8

Impact of varying the progressivity of the capital and the bequest tax in a model without land (cf. Sect. 3.1.3). If only the top quintile is taxed, both tax instruments have a higher potential to reduce wealth inequality compared to scenarios in which a greater fraction, or all households are taxed. A capital tax on the top quintile has a higher potential to reduce inequality and a less negative impact on output than a bequest tax on the top $20\%$

Table 6 Gini coefficients of the wealth distributions of those countries for which OECD (2015) reports data on their respective wealth distribution

Kaldor–Hicks criterion

Even though we find that recycling all public revenues to the young as lump-sum transfers enhances output and reduced inequality, a Pareto improvement is not possible with such a transfer scheme. However, we find that at least there are cases in which the Kaldor–Hicks criterion is fulfilled. Consider, for instance, the case in which all land rents are skimmed off and redistributed to the young ($\tau _L = 1$, $\delta = 0$) shown in Figs. 9 and 10. Absent any additional transfer mechanism between winners and losers, generations belonging to the top wealth quintile $i=5$ and possibly all households belonging to the first old generation suffer under the tax. Whether the first old generation suffers under the tax depends on the assumption made about its exogenously fixed level of consumption when young.

Now we introduce a mechanism that allows intertemporal transfers between households. Instead of the lump-sum transfers from public revenues $g_t$, young and old households may now receive a transfer or have to pay a lump-sum tax X. Their budget equations thus are

$$\begin{aligned} c^y_{i,t}+ s_{i,t}&= w_t + b_{i,t}(1-\tau _B) + X^y_{i,t}\\ c^o_{i,t} + b_{i,t}&= (1+R_{t}(1-\tau _K))k_{i,t}^s + l_{i,t}(p_{t}+q_{t}(1-\tau _L)) + X^o_{i,t}. \end{aligned}$$

Further, we assume that funds can be shifted over time via banking and borrowing at the market interest rate R. Then, for the total volume of the transfers, it has to hold that

$$\begin{aligned} \sum _t \frac{g_t}{\Pi _{s=1}^t (1+R_s)}&\ge \frac{1}{N}\sum _{i,t} \frac{X^y_{i,t} + X^o_{i,t}}{\Pi _{s=1}^t (1+R_s)}. \end{aligned}$$

Our numerical experiments confirm that there are feasible combinations of $\{X^y_{i,t}, X^o_{i,t}\}_{i=1,\ldots ,N,\ t=1,\ldots ,T}$ such that the winners of the 100% land rent tax can compensate the losers, i.e., that

$$\begin{aligned} u_{i,t}|_{\tau _L=1}&\ge u_{i,t}|_{\tau _L=0} \quad \forall i,t. \end{aligned}$$

Fig. 9

When land rents are taxed at 100% and recycled as lump-sum transfers to the young, the richest households bear the burden. Their utility under taxation is less than without taxation, i.e., $u|_{\tau _L=1} - u|_{\tau _L=0} < 0$. All other households benefit from the policy

Fig. 10

For certain parameters, not only the rich households, but also the households belonging to the first old generation bear the burden of the tax reform