The Taxation of Norway’s Richest¹

Aaberge et al. (2020) studied inequality in income and wealth in the tax system in Norway in the period 2001-2018. They use more complete measures for income and wealth that are omitted in official statistics by including personal owners’ shares of retained earnings in private companies.² Based on detailed Norwegian register data on company ownership, they assign the income of the companies and corporate taxes paid to the owners. Their measure of market income supplements the official statistics by considering income that is earned in companies but not paid out as dividends. This is important because Bjerksund and Schjelderup (2021) show that the Norwegian system of shareholder taxation creates a strong incentive to lock capital in companies.³ This is due to the fact that capital gains from shares are tax-exempt at the corporate level, while dividend payouts to personal shareholders are taxed.⁴ The lock-in effect stems from that the value of deferring the personal tax on dividend exceeds the after-tax value of dividend payout to the personal shareholder. This lock-in effect has led to a substantial build-up of capital in Norwegian companies that affects the burden of tax among Norwegian taxpayers (see NOU (2022, p. 203)).

In two studies, Aaberge et al. (2020, 2021) show that retained earnings in the corporate sector affect the progressivity of the tax system over time. Aaberge et al. (2020) found that for the period 2004–2018, the richest 1 percent in Norway paid, on average, around 22 percent in tax, while the 9 percent who are among the richest 90 to 99 percent paid an average of 33 percent tax on their income. These findings show that the Norwegian tax system is regressive at the top of the income distribution. This is underlined by the fact that among the top 0.1 percent (about 3,700 persons) the lowest effective tax rate in the period 2004–2018 was 9 percent, while the highest was 17 percent.

In accordance with the findings of Aaberge et al. (2020), we present an example of an investor who earns exclusively capital income through firm ownership.⁵ This serves the dual purpose of reconciling Aaberge et al.’s findings and examining why the Norwegian tax system is regressive at higher income levels through a detailed analysis of tax payments. Our example offers a detailed analysis of tax contributions, thereby shedding light on the reasons behind the tax system’s regressive nature at higher income levels.

Our example shows that from 2004–2018, the average effective tax rate for Norwegian investors ranged from 14–21 percent, varying based on their efforts to minimize tax payments. The lower end of this interval and the notably low rate of 14 percent are attributable to investors who have strategically minimized their taxes. They leveraged the 2006 tax reform to avoid dividend taxes, held only unlisted shares to reduce wealth tax, and contributed to taxes primarily through the corporate taxes paid by their firms. The favorable assessed value of unlisted shares, the low effective average corporate tax, and the fact that their income stems from capital gains are key reasons why the tax system is regressive at the upper end of the income distribution. Avoiding the dividend tax does not matter much for our results in the main scenario.

Figure 1 presents findings from Aaberge (2021), in which the population is divided into deciles. The figure illustrates that the top 1 percent of the wealthiest individuals own the majority of stocks in the economy and that the top 0.1 percent earn almost all their income from stock holdings while paying very little in taxes. The regressivity of the tax system is evident, as the wealthiest 1 percent and 0.1 percent pay less in taxes compared to other groups in the economy.

Figure 1:

The composition of income among different groups in Norway ranked according to income category, average 2006 – 2018 (source: Aaberge et al. (2021)).

Y-axis percentage. X-axis deciles.

The blue pillars represent income from company ownership, the red pillars represent wage income, the white pillars represent income from housing, the green pillars are municipality services, and the brown pillars represent tax payments.

Our study relates to a strand of literature on effective tax rates across income groups. Piketty et al. (2018) calculate inequality statistics for the U.S. by developing distributional national accounts. They include retained earnings as part of the national income on an accrual basis within their analysis. A particularly notable discovery is that a substantial portion of the increase in top incomes since the late 1990s can be attributed to capital gains from equities and bonds. Utilizing U.S. tax returns and data from the Fortune 400, Yagan (2023) finds that the effective income tax rate for the 400 wealthiest American households is only 9.6 percent. In a similar vein, forthcoming research from France by Bozio et al. and studies from the Netherlands by Bruil et al. (2022) reveal tax regressivity at the upper end of the income distribution. The marked disparity between statutory and effective tax rates in these studies is primarily due to the concentration of control over retained earnings and unrealized capital gains at the upper end of the income distribution.

There is also a body of literature that compares statutory tax rates (the rates set by law) with effective tax rates (the actual rates paid after deductions and credits). One influential work in this area is De Mooij and Devereux (2011), who compare effective tax rates on equity and debt financing in different countries, highlighting the bias toward debt financing in many tax systems. Related is Markle and Shackelford (2012), who compare effective tax rates across countries and over time, finding significant international variation.⁶ These studies, and others like them, contribute to our understanding of how tax systems operate in practice and how they can influence economic behavior and cross-border investment.

First, we describe the changes in the Norwegian tax system during the period 2004–2018. We then set up a model of a private investor who has a holding company that owns a daughter company. Norwegian investors organize themselves this way because the participation exemption allows them to receive dividends and capital gains from shares tax-free across companies without paying tax. We study two scenarios: The main scenario depicts an investor who has strategically planned to minimize tax payments, and the secondary scenario describes an investor who passively pays taxes without any effort to reduce tax obligations. The purpose of these scenarios is to identify the range within which the effective tax rate for the wealthiest individuals falls.

The Norwegian income tax system has, since 1992, been based on the principle that all taxable incomes should be included in a common income base with a flat tax rate (ordinary income). Considerations of fairness have been addressed through progressive taxation of gross labor and pension incomes, while efficiency concerns have guided the lower and flat tax rates on both corporate and individual capital incomes.

During the period 1992–2006, all forms of capital income were taxed at a rate of 28 percent, but there was no tax on dividends. Incomes earned in the corporate sector were taxed at the corporate level but, in practice, not at the shareholders’ level upon the distribution of stock dividends or the realization of stock gains. This was due to the so-called compensation method for stock dividends and the RISK method for stock gains, both designed to prevent double taxation of the corporation and the shareholder, as well as chain taxation.⁷

Dividend taxes.

Estimating the effective tax on rich investors based on the tax code is a difficult exercise, and the case of dividend taxes is particularly challenging.

Prior to 2006, owners with an ownership stake exceeding two-thirds were taxed under the split model, which was a sort of “substitute” for the dividend tax. Under the split model, a return on the labor effort of the active owner was imputed and taxed as labor income. The taxation of imputed labor income was done independently of whether this income was paid as wages, dividends, or retained in the firm. The split model provided a strong incentive for active owners to shift income from labor to capital income by reducing their ownership stake below two-thirds, and by 2006, very few companies had owners with an ownership stake of two-thirds.

The introduction of the 2006 Norwegian shareholder income tax was announced in advance and therefore allowed shareholders to take measures to minimize tax. The shareholder model increased top marginal tax rates on individual dividend income from zero to 28 percent and, thus, provided strong incentives to pay out extraordinary dividends prior to 2006, which is evident from Figure 2. Some of the richest investors in Norway received billions in untaxed dividends and subsequently reinvested these payouts as equity in their firms.⁸ Figure 2 shows dividend payouts over time. The decision to introduce a dividend tax was announced in the spring of 2004. The knowledge that one would have to pay tax on dividends of 28 percent from the income year 2006 likely contributed strongly to the significant increase in tax-free paid-out dividends in 2005 and the large drop from 2005–2006. Similarly, the increase in dividends from 2014–2015 is probably also an adjustment to a notified increase in the dividend tax: The tax rate for dividends to personal shareholders was raised from 27 to 31.7 percent over four years, first to 28.75 percent in 2016. It should also be noted that a large share of dividends is accrued by the top 1 percent.

Figure 2:

Dividends to shareholders. Green line total dividends. Black line dividends to top 1% richest. Source: Aaberge et al. (2020).

Figure 2 also reveals another interesting fact. Before 2006, almost all dividends were received by the top 1 percent. After 2006, a smaller share of dividends was received by this group. A likely explanation is that the top 1 percent reinvested dividends received prior to the 2006 reform back into their firms as equity and used equity withdrawals as a substitute for dividends.⁹ In contrast, other income groups needed to receive dividends to finance their consumption.

The shareholder model was combined with participation exemption rules (“Fritaksmetoden”). These rules are present today and imply that corporate shareholders are generally exempt from tax on dividends received, on capital gains from qualifying shares, and on derivatives where the underlying object is qualifying shares. Correspondingly, corporate losses on qualifying shares are non-deductible. All operating expenses related to exempt income from shares (e.g., management costs) are fully tax-deductible. Dividends received only fall under the participation exemption for 97 percent of the received value, with the remaining 3 percent taxable at the standard corporate rate of 22 percent, providing for an effective tax rate of 0.66 percent. Dividend distributions within a tax group (where the ultimate parent company directly or indirectly owns more than 90 percent of the shares and voting rights) are, however, fully tax-exempt, reflecting that an alternative in many cases would be a tax-neutral group contribution.

The ability to write down equity serves as a viable alternative to receiving dividends and raises the question of whether the richest 0.1 percent paid any dividend taxes during the period 2004–2018. We will revisit this matter later.

Figure 3 shows how nominal corporate and dividend taxes have varied in the period 2004–2018.

Figure 3:

Headline corporate tax rates and marginal dividend tax rates (percentage).

Corporate taxes.

The study by Aaberge et al. (2020) spans the period 2001–2018. In this period, corporate and dividend taxes varied substantially. Aaberge et al. (2020) report average tax rates for the richest top 0.1 percent for the period 2004–2018, and we will therefore focus on this period as well. The weighted average of corporate tax rates during this period is approximately 27 percent and the average for marginal dividend taxes is approximately 24.5 percent, c.f. Appendix. Figure 2 shows how dividend and corporate taxes have changed over time.

The real effective corporate tax rate is taxes paid as a percentage of gross corporate income. The tax rate paid by the company may deviate from the nominal rate because taxable profit does not correspond to real economic profit.¹⁰ If true economic profit is NOK 100, the taxable profit is NOK 80, and the tax rate is 27 percent, taxes paid is NOK 21.6, and the effective tax rate is 21.6 percent (i.e., 21.6/100). It is likely that the real effective corporate tax rate differs from the nominal headline rate of 27 percent.¹¹

One can use empirical data to estimate the empirical AETR by dividing a corporation’s total taxes paid by its total pre-tax taxable profits. This rate is considered more informative than the statutory or nominal corporate tax rate because it reflects the real-world tax impact on a company’s profits. The formula for calculating the Average Effective Corporate Tax Rate (AETR) is as follows: 1AETR= Corporate taxes paid Annual profit + corporate taxes paid .

The empirical AETR is a useful metric for policymakers, researchers, and analysts to understand the actual tax burden on corporations, as it reflects the impact of the entire tax code, including various provisions that can affect a company’s taxable income. It provides a more nuanced perspective on corporate taxation than the statutory rate alone.¹² We estimated the AETR using the Norwegian Corporate Accounts database, organized and documented by Mjøs and Selle (2022). The Norwegian corporate accounts database is a comprehensive micro-level database of fundamental financial and corporate information on Norwegian companies, financed by the Centre for Applied Research at NHH (SNF). The database consists of more than 6 million firm-year observations for legal entities and 140,000 corporate groups in the period 1992–2020. The database is organized around a unique organizational number issued and administered by the Brønnøysund Register Centre, which also allows for merging with other data sources. The database contains data from several sources. Complete financial records, including investments, have been collected from Bisnode D&B Norway AS. The Brønnøysund Register Centre also provides a variety of corporate information, e.g., company name and ID, legal incorporation form, industry codes and business sectors, date of establishment, business and legal addresses, and number of employees. Ownership information is from the Shareholder Registry, maintained by the Norwegian Tax Authorities. These records include information on all shareholders in Norwegian limited liability firms.

In our analysis, we study unconsolidated financial records of active limited liability firms in the period 2004–2018 in all sectors except oil and finance. To prevent our findings from being influenced by outliers, we implemented a few gentle sample restrictions. First, we exclude observations with very high levels of profitability or losses. We set the cut-off at a return on assets exceeding plus or minus 100% (163,054 observations). Second, we exclude observations with very high levels of debt. We set the cut-off at observations where total debt is more than three times the total assets (23,405 observations). We are then left with 2,341,564 firm-year observations.

Figure 4 shows the AETR for the period 2004–2018. The AETR varies from 10.5 percent in 2018 to 16.6 percent in 2004 and reflects changes in the headline tax rate and the tax base.¹³ The average AETR for the whole period is 13.1 percent, which is the rate we will use in what follows.

Figure 4:

Average effective corporate tax rates (AETR) (percentage).

A common way of organizing ownership among the richest in Norway is to establish a holding company that owns shares in other companies. This way, dividend payouts can be accumulated tax-free in the holding company due to participation exemption. To capture the effective tax rate of a wealthy investor, we set up a model that reflects this arrangement. The investor (owner) we study is the sole owner of a holding company that wholly owns a subsidiary. We have simplified this structure down to two companies, as shown in Figure 5.¹⁴

Figure 5:

The ownership structure.

Jordà et al. (2019) estimate total annual returns for equity, housing, bonds, and bills across 16 advanced economies from 1870 to 2015. For Norway, post-1980, the return on equity is 12.22 percent. For the full sample of 16 countries, the unweighted average return on equity post-1980 is 10.78 percent, while the weighted average return is 9.08 percent. We will use the return on Norwegian equity in our analysis, but it is important to note that using any of these rates has a very small effect on our findings.

We shall assume throughout that the subsidiary has a taxable income stream each year of NOK 100. Using 12.22 percent as the discount rate, the discounted value of the company’s income stream of NOK 100 annually values the company at 2NOK10012.22%≅NOK818.

The market value of stocks that are listed is the price of the share on January 1 in the tax year on the stock exchange. In our main scenario we assume that the company is a non-listed company because the studies by Aaberge et al. (2020, 2021) show that the presence of unlisted stocks is significant among the very wealthiest individuals.

The wealth value of unlisted stocks is assessed as the stock’s share of the book value of the company’s equity. Andresen and Bø (2022) find that the valuation discount is 65 percent of market value for non-listed shares. Myklebust and Øverland (2022) estimate this discount at 68 percent. We use the lower estimate of the discount, which is 65 percent of market value, so that the assessed value for the wealth tax is 35 percent of the market value. Consequently, the wealth tax assessed value before any rebate is NOK 818 · 35% = NOK 286.

The wealth tax has varied from 1.1 percent at the start of the measurement period to 0.85 percent in 2018. The average headline wealth tax rate for the period is approximately 1 percent. The rebate on the assessed value of stocks (referred to as “formuesverdi”) has ranged from 35 percent to 0 percent of the assessed stock value. The average in the period is 6.7 percent on the wealth tax base, c.f. Appendix. Hence, starting with the value NOK 286, there is then a rebate on this value of 6.7 percent, which leaves the tax base of the wealth tax at NOK 818 · 35% · (100% – 6.7%) = NOK 267.2

The wealth tax that falls on the investor if he holds only unlisted stocks is calculated as net wealth minus debt multiplied by the wealth tax rate of 1 percent.

In calculating the debt of the investor, we have utilized the Norwegian microdatabase (www.microdata.no) to estimate the investor’s debt level. This database allows us to estimate debt based on income levels. For an individual earning NOK 100, the debt would amount to 134 percent of their gross income, equating to NOK 134 in this example. While the actual debt level for the top 0.1 percent wealthiest individuals could be lower or higher, this figure represents our best estimate. In Section 6, we will relax the assumption that the investor has debt and assume he has no debt. Using NOK 134 as his debt, the wealth tax basis if the investor holds unlisted shares are 3(NOK 818·35%·(100%−6.7%)− NOK 134)·1%≅ NOK 1.33

The effective wealth tax rate, calculated as a percentage of a gross income of NOK 100, is 1.33 percent. Aaberge et al. (2020) found that the wealth tax varied between 1 and 2 percent for the wealthiest 5 percent. The rebates on unlisted shares reduce the market value of NOK 818 to NOK 267, and after subtracting the debt of NOK 134, the wealth tax base is NOK 133, which explains why the wealth tax the investor must pay is so low.

If the investor holds listed stocks, the basis for the wealth tax is market value. The market value of the shares in our example is NOK 818. The tax rebate on average is 6.7 percent in the period for listed shares so the wealth tax falling on listed shares in our example is given by 4(NOK 818·(100%−6.7%)− NOK 134)·1%≅ NOK 6.3

It is worth noting the significant discrepancy in valuation between listed and unlisted stocks, which plays a crucial role in calculating the wealth tax.

The real effective corporate tax rate is the subsidiaries AETR, which on average for the period 2004–2018 is 13.1 percent, c.f. Appendix. The subsidiary has a taxable profit of NOK 100, and Table 1 shows the financial accounts of the subsidiary.

After paying the corporate tax, the subsidiary pays out the residual surplus of NOK 86.9 in dividends to Ltd. Holding. We assume that the holding company exclusively receives dividends from its subsidiary and has no other income. Under the participation exemption rules, corporate shareholders are generally exempt from tax on dividends received, on capital gains from qualifying shares, and on derivatives where the underlying object is qualifying shares. Correspondingly, corporate losses on qualifying shares are non-deductible. Hence, dividends are taxed only when they are paid out to personal shareholders.

Our main scenario involves an investor who has extensively strategized to minimize tax payments. Anticipating the 2006 tax reform, he secured extraordinarily high dividends that were subsequently reinvested as equity in his firms. After the introduction of the tax reform in 2006 and the dividend tax, he opted not to draw dividends and instead reduced his injected equity. Moreover, he exclusively holds unlisted stocks to minimize wealth tax obligations.

There are other reasons why tax planners, such as our investor, avoid taxable dividends. First, under the shareholder model, only returns exceeding the riskfree interest rate are subject to dividend tax, enabling wealthy investors to receive substantial dividends taxfree. Second, the 2006 tax reform introduced the shareholder model in conjunction with participation exemption rules (“Fritaksmetoden”), exempting corporate shareholders from taxes on dividends received, capital gains from qualifying shares, and related derivatives. These provisions made it profitable for owners to reinvest capital into their firms for further investment and savings rather than distributing it as dividends.¹⁵

To summarize, in our main scenario, the investor has no labor income, exclusively owns unlisted shares, and avoids dividend taxes. Consequently, he only pays the corporate tax and the wealth tax. He pays the wealth tax using tax-exempt dividends and/or by writing down equity.¹⁶ Ideally, we should model this aspect, but the amounts involved are so minimal that they have a negligible impact on our results.

Since the investor does not need to receive dividends from Ltd. Holding, the financial account of Ltd. Holding is given in Table 2. Dividend and wealth tax payments (c.f. equation (3)) by the investor are given in Table 3.

Based on Tables 1–3, we calculate the average tax burden on the owner. We use the same methodology as Aaberge et al. (2020), taking into account both the taxes paid by the investor through their ownership and the gross income of the firms they own when calculating the effective tax rate. The gross income of the owner is equal to the profit of NOK 100 in the subsidiary. Total taxes paid include the corporate tax (NOK 13.1) and the wealth tax (NOK 1.33). Our estimate of the effective tax of the 0.1 percent richest in Norway is then 5NOK 13.1+ NOK 1.33 NOK 100≅14.4%

An effective rate of 14.4 percent, as an average over the entire period 2004–2018, aligns closely with the findings in Aaberge et al. (2020) for the richest 0.1 percent in Norway. The top 0.1 percent had an average tax burden between 9 and 17 percent during this period according to Aaberge et al. (2020). Moreover, the 0.9 percent who are among the richest 99–99.9 percent paid on average around 22 percent in tax in the same period, while the 9 percent who are among the richest 90–99 percent paid an average effective tax of 33 percent.

The average effective dividend tax between 2004–2018 is 24.5 percent, c.f. Appendix. Note that even if we abandoned our assumption that the investor does not need to receive dividends to pay the wealth tax, the calculated effective tax rate would not increase significantly. Relaxing this assumption implies that the investor must receive NOK 1.76 in dividends and pay NOK 0.43 in dividend taxes.¹⁷ This would increase the effective tax rate to 14.9 percent.¹⁸

The secondary scenario describes an investor who passively pays taxes without any effort to reduce tax obligations. Thus, the investor must receive dividends to pay the wealth tax, and he holds only listed shares. To cover the cost of the wealth tax, which is NOK 6.3 (see equation (4)), the investor must withdraw sufficient dividends from his holding company. The average effective dividend tax between 2004–2018 is 24.5 percent so the investor needs to receive dividends in the order of 6NOK 6.3100%−24.5%≅ NOK 8.3

The books of the holding company and the investor are then given by Tables 4 and 5.

As before, the gross income of the owner is equal to the profit of NOK 100 in the subsidiary. Total taxes paid are the corporate tax (NOK 13.1), the wealth tax (NOK 6.3), and the dividend tax (NOK 2.0), and the effective rate of tax is 7NOK 13.1+ NOK 6.3+ NOK 2.0 NOK 100≅21.4%.

Based on our revised assumptions, the effective rate of tax is 21.4 percent over the period 2004–2018. It is interesting to note that the average single worker faced a net average tax rate of 27.3 percent in 2022, compared with the OECD average of 24.6 percent.¹⁹ Aaberge et al. (2020) find that, in our translation: “During the period 2004–2018, the wealthiest 1 percent on average paid about 220 kroner in tax per 1,000 kroner in income, while the next 9 percent of the richest, from the 90th to 99th percentiles, on average paid 330 kroner in tax per 1,000 kroner in income." Our number of 21.4 is almost at par with what Aaberge et al. (2020) found for the top 1 percent.

We should stress that our example assumes that the richest do not have any personal (wage) income. This assumption underestimates taxes paid. Aaberge et al. (2020, 2021) find that company income accounts for over 90 percent of the gross income of the richest 0.1 percent, but that they do have some labor income. Our assumption can be defended on the grounds that any labor income is so small that it does not affect our interval in a substantial way.

The assumption that the investor has a debt of NOK 134 could be disputed. If we assume that the investor has no debt, it follows from equations (3) and (4) that the wealth tax base increases by NOK 134 and that the wealth tax increases by NOK 1.34 in both scenarios.

In the main scenario, the investor holds unlisted stocks. In case of no debt, the effective tax rate is 8NOK 13.1+(NOK 1.33+ NOK 1.34) NOK 100= NOK 13.1+ NOK 2.67 NOK 100≅15.8%, as compared to 14.4% in equation (5) above.

In the secondary scenario, the investor holds listed stocks and must receive dividends to pay the wealth tax. In case of no debt, the investor needs to receive dividends in the order of 9NOK 6.3+ NOK 1.34100%−24.5%= NOK 7.6100%−24.5%≅ NOK 10.1, c.f. equation (6). The dividend payment (NOK 10.1) will then cover the wealth tax (NOK 7.6) and the dividend tax (NOK 10.1 · 24.5% = NOK 2.5). Total taxes paid are the corporate tax (NOK 13.1), the wealth tax, and the dividend tax. The effective tax rate of tax in this case is 10NOK 13.1+ NOK 7.6+ NOK 2.5 NOK 100≅23.2%, as compared to 21.4% in equation (7).

To sum up, we find that throughout the period from 2004 to 2018, the average effective tax rate for Norwegian investors without debt ranged from 16 percent to 23 percent, varying based on their efforts to minimize tax payments.

In our example, we have used the standard definition of income found in economic literature and official statistics: income is the money that a person or business receives in exchange for selling a good, performing a service, or as a return on invested capital. According to this definition, retained earnings in companies are part of the return on invested capital and should therefore be included in the measurement of income when calculating the effective tax rate for the investor in our example.

Critics of including returns to capital argue that retained earnings in a company cannot be equated with income that individuals receive as dividends from the same company or as wages paid for work performed. They argue that for retained earnings to be considered as personal income, a ’potential dividend tax’ should be deducted from those earnings. A counterargument is that the income a person controls can either be consumed or saved, regardless of whether the income comes from labor or capital. The retained surplus in companies can be regarded as savings. It is therefore reasonable to assume that these saved retained surpluses are invested in various assets, contributing to the wealth increase of the owners. If the surplus is retained in the company and not paid out as taxable dividends, the saved tax also becomes part of the income that can be invested in various assets. Thus, both the retained surpluses and the saved tax will contribute to wealth accumulation for the owners of the companies. However, whether and how this increase in wealth will be taxed, nobody knows. It will also depend on possible future changes in the tax system. This is the reason why studies like Aaberge et al. (2020, 2021) do not deduct a hypothetical future tax in their calculations.

Another objection to including retained corporate income is that many owners of large, listed companies have limited influence over the amount paid out in dividends. However, if profits are retained in such companies, it is reasonable to expect this to be reflected in the companies’ market values. These values can be realized by the owners whenever they wish by selling their shares. Interestingly, Aaberge et al. (2020, 2021) demonstrate that large retained surpluses are primarily found in non-listed (holding) companies, where the owners have significant influence over dividend decisions.

In our analysis we have tried to reconcile the findings in Aaberge et al. (2020) through an example. Our main scenario is of an investor who has done his utmost to save taxes. He entirely avoids the dividend tax and only holds unlisted stocks to save wealth taxes. He pays the wealth tax by reducing equity in his firm or by receiving tax free dividends. In this case the effective tax is 14.4 percent. If we relax the assumption that the investor avoids the dividend tax, this increase or estimate in the main scenario to 14.9 percent. Thus, the assumption that the investor avoids the dividend tax is not important for our results. Our secondary scenario is of a passive taxpayer who holds listed shares and must finance the payment of the wealth tax by withdrawing dividends. His effective tax rate is 21.4 percent.

Our example shows that throughout the period from 2004 to 2018, the average effective tax rate for Norwegian investors ranged from 14 to 21 percent, varying based on their efforts to minimize tax payments. If we relate this interval to the effective rate of tax for lower income groups in Norway, the example confirms that the tax system is regressive at the top even at the top end of this interval (see Aaberge et al. (2020, 2021)). Thus, the Norwegian tax system demonstrates a weak capacity for income redistribution at the top of the income distribution. This weakness stems from the fact that the wealthiest earn almost no wage income, with the majority of their income deriving from capital.

The primary reason for the low effective tax rate in our main scenario is the significantly low tax-assessed value of unlisted stocks and the low corporate AETR. Holding unlisted shares allows the investor to reduce the wealth tax from NOK 6.3 million for listed stocks to NOK 1.3 million for unlisted stocks, thereby reducing the wealth tax burden by 78.8 percent. Consequently, the low valuation basis for stocks in calculating the wealth tax renders it an ineffective redistributive tool.

Finally, it should be noted that Aaberge et al. (2020, 2021) are empirical studies where income and taxes are gathered from various data sources to analyze the effective tax rates of different strata of the population. Our study is an example of an investor who has organized his business the way most wealthy Norwegians do. Based on this it is quite striking that our findings are so closely aligned with their results. One can, therefore, argue that the low effective tax rates for the top 1 percent and the top 0.1 percent wealthiest, is a result of changes in the Norwegian tax system that largely has benefitted the richest part of the population.