Inheritance Taxation in Sweden, 1885–2004: The Role of Ideology, Family Firms, and Tax Avoidance

This article studies the evolution of Swedish inheritance taxation since the late nineteenth century to its abolition in 2004. The contribution of this article is twofold. First, the annual effective inheritance tax rates are computed for different sizes of bequests and asset types, accounting for all relevant exemptions, deductions, and valuation discounts. Second, an attempt is made to explain changes in inheritance taxation over time. Ideology appears to be the main driver of the sharp tax increases of the 1930s to the 1960s. Wartime economies with higher pressures on the people induced politicians to raise inheritance taxes on the wealthy, primarily during the First World War. Increased opportunities for tax planning for the wealthy are also documented, most notably a series of tax cuts on inherited family firms in the 1970s. This rise in avoidance opportunities for the rich, while middle&#8208;class heirs faced growing inheritance tax rates, undermined the legitimacy of the tax and led to its repeal.

T he evolution of inheritance, gift, and estate taxation across different economic systems engenders important questions regarding the trade off between egalitarian ambitions and incentive effects in the welfare state. Inheritance taxes are among the most direct fiscal instruments for equal opportunity in every new generation. Simultaneously, however, they may dampen incentives to accumulate wealth and induce tax evasion and avoidance behaviours. Ultimately, whether such taxes positively or negatively contribute to societal development over the short and long run thus becomes an empirical question.
This study offers two contributions to the literature on inheritance taxation. Most studies of long-run trends in inheritance taxation base their analysis on statutory top marginal tax rates, that is, the highest possible marginal rate paid by heirs regardless of the amount of inheritance required to reach that rate. In contrast, this article presents a new long-run series of effective average inheritance tax rates computed on the basis of the full spectrum of institutional factors affecting the actual final tax payment. Furthermore, these rates are presented for differently sized inheritances, from an average middle-class heir to an extreme upper-class heir, and for different types of inherited assets: family-firm equity and non-firm wealth. Our series are annual, they span from 1885 to 2004, and cover a period from the early stages of industrialization to the near present, that is, until 2004, when inheritance tax was abolished.
A second contribution of our study is an analysis of the main driving forces that underlay the changes in inheritance taxation in Sweden, from the increases during the interwar and immediate postwar periods to the gradual dismantling beginning in the 1970s and the final repeal in 2004. We propose four main explanations for the observed patterns. First, the expansion of the tax in the 1930s and late 1940s appears to be primarily related to an ideology of redistribution and the fact that Social-Democratic governments were in power from 1932. However, the Social Democrats did not form a one-party government based on solid parliamentary support until after the end of the Second World War. We argue that the delay in this political development effectively curbed the extent of inheritance tax increases. Second, the role of mass mobilization during the two world wars, which Scheve and Stasavage have identified as an important determinant for the significant increases in inheritance tax rates in many countries, also appears to have mattered in Sweden and thus complements the ideology channel. 1 Third, our estimates indicate an increasing degree of tax avoidance and evasion during the postwar era, a problem noted by contemporary legislators who acted to halt additional tax increases and who eventually began reducing effective tax levels from the 1970s onwards. Fourth, the demise of the Swedish inheritance tax seems to result from a loss of legitimacy among broad layers of the population in the latter part of the twentieth century. Soaring house prices in combination with bracket creep during the 1990s doubled the inheritance tax rate, which was paid by almost one-third of all adult heirs. Meanwhile, because of new legislation in the late 1990s, the tax burden at the top end of the distribution was effectively reduced. These events undermined the legitimacy of the inheritance tax for Swedish taxpayers and eventually caused its abolition.
Research on trends in Swedish inheritance taxation is of interest beyond the specific historical experiences in Sweden. The recent work on the long-run evolution of inheritance flows in France by Piketty and by Piketty, Postel-Vinay, and Rosenthal and the related study of inheritance flows in Sweden from the early nineteenth century to the present by Ohlsson, Roine, and Waldenström have shown that the relative importance of inheritance in aggregrate terms has changed dramatically over time. 2 These studies have also shown that this change has largely resulted from macroeconomic factors such as the relationship between private wealth accumulation and income growth.
What has received less attention is the role of institutional developments in this process, particularly the evolution of a political democracy and the taxation of inheritance and wealth. In a historical overview of the political debates surrounding inheritance taxation in Germany, France, and the US, Beckert emphasizes the importance of studying national institutions to gain a full understanding of the evolution of inheritance taxation. 3 Over the period studied, Sweden developed the world's most extensive welfare state with a strong egalitarian emphasis. 4 Examining inheritance taxation from a historical perspective is crucial not only to understand the achievements of the Swedish welfare state but also to gain further insight into the society in which it gained popular support. According to some scholars, the Swedish tax system became increasingly hostile towards entrepreneurship and business ownership during the postwar period. 5 If these scholars are correct, what was the role of inheritance taxation in this development, particularly with regard to the transfer of family firms to the next generation?
Previous research has not extensively studied the endogeneity of inheritance flows regarding inheritance taxation. However, a recent study by Piketty and Saez shows that the elasticity of bequests to the tax rate is related to the use of standard tax avoidance technologies, the concentration of bequests, and the importance of small inheritances in society. 6 Although addressing all these dimensions is beyond the scope of our analysis, we specifically investigate the extent of tax avoidance opportunities and their potential effect on observed inheritance flows in the case of Sweden.
Section I provides an overview of the rules and tax rates that determined inheritance taxation in Sweden over the entire period of study. Section II then presents a calculation of the effective inheritance tax rates for different sizes of estates containing either personal net wealth or family business equity. Section III presents additional facts and speculation regarding the main determinants of the evolution of inheritance taxation in Sweden, and section IV concludes. I This section presents the basic principles, including tax schedules, deductions, exemptions, and valuation rules, that have determined Swedish inheritance, gift, and estate taxation since the emergence of modern inheritance taxation in 1884 until its final abolition in 2004. 7 This information is required to calculate the effective inheritance tax rates presented and analysed in the remainder of the study.
Taxation of a deceased person's assets in Sweden has predominantly been implemented in the form of an inheritance tax on the legacies received by the estate's beneficiaries. Internationally, this method is the most common form of taxation of intergenerational transfers, and it differs from estate taxation, where the wealth of the deceased is taxed. 8 The starting point for calculating inheritance tax is the remainder of a deceased person's estate after outstanding debts and, if the deceased was married, the spouse's right to marital property are settled. The remainder is allocated to the heirs and beneficiaries under the will, and, as a final step, the inheritance tax is calculated for each heir. Assets included in the taxable estate are real and financial assets, including consumer durables and most private insurances, remaining after debts. The tax-exempt spousal property removed from 5 Henrekson and Jakobsson, 'Schumpeter';Lindbeck,'Swedish experiment'. 6 Piketty and Saez,'Theory'. 7 Various types of duties and fees on estates, inheritances, and wills existed previously but only for small and specific parts of the tax base and population strata. See Du Rietz, Henrekson, and Waldenström, 'Swedish inheritance', for an exhaustive description of these rules and regulations. 8 For more extensive overviews of inheritance, gift, and estate taxes, see, for example, Gale and Slemrod, 'Rethinking'; Boadway, Chamberlain, and Emmerson, 'Taxation of wealth'; and Kopczuk, 'Taxation of intergenerational transfers'. the taxable estate has typically amounted to half of the estate. Since 1960, it has amounted to at least four times the basic amount allowed adjusted for price inflation. 9 The division of the estate into taxable inheritance lots is based on legal rules of inheritance order across different classes of heirs and stipulations in the deceased's will, if one exists. For example, if there are three children, the estate is divided into three equal parts unless an existing will stipulates differently. If an heir abstains from his or her inheritance, the individual heir's share of the estate is passed on to his or her children.
Gift taxation is an integral part of any inheritance tax system. If every gift were considered independent of earlier acquisitions, large tax gains could be accomplished simply by dividing gifts into smaller instalments, or inter vivos gifts (literally meaning gifts between the living), that are allocated over time. To counteract tax avoidance, specific summation rules were introduced early on in the Swedish inheritance and gift tax ordinance, stipulating that gifts and bequests from the same donor should be added to inheritance lots and taxed jointly.
Valuation of assets and liabilities in the tax base comprise a central component of inheritance taxation. The starting point for the valuation of the assets and liabilities of estate inventories is their market value at the time of death of the deceased. However, several special valuation rules have been applied on different asset classes in the Swedish inheritance tax code, and several important changes in the valuation principles that determine the final tax burden have been made over time.
Real estate was typically recorded at its tax-assessed value in the year preceding death. Other personal property was taxed at market value, and businesses were valued at their sales value estimated by trustees. However, some asset classes were listed at only a fraction of their market value. For example, shares registered on the main listings at the Stockholm Stock Exchange were listed at 80 per cent of their full market value from 1997 to 2004, 75 per cent from 1978 to 1996, and 100 per cent before 1978. Unlisted shares were assessed at only 30 per cent of their quoted or book value beginning in 1978. Forest holdings were listed at half their market value throughout the period under study. Small firm inventories and stock-in-trade have also occasionally been valued below market prices (see more below).
Insurance policies represent a specific and relatively problematic asset class in inheritance taxation. If a deceased person leaves behind insurance without beneficiaries, the value of the insurance, or the insurance disbursements, is simply included in the estate inventory. The same principle normally applies to insurance possessed by a surviving spouse. However, insurance with beneficiaries, who are actually included in most insurance contracts, is typically tax exempt, complying with marriage codes. 10 Some variations nevertheless exist in the rules governing 9 The price-adjusted amount is changed when the general price level changes. Calculations are based on the change in the Consumer Price Index and established for the entire calendar year. Many transfer payments, tax rates, entitlements, and so forth, are determined by the price-adjusted amount. In 2004, the price-adjusted amount was SEK 39,300, and an average worker's annual salary was SEK 262,200. The non-taxed spouse's marital property for that year thus amounted to 60% of the average annual salary. 10 Englund, Beskattning av arv och gåva.
the treatment of insurance acquisitions both over time and across certain insurance types. 11 Certain forms of business assets received substantial tax and valuation relief in the latter part of the twentieth century. This relief represents an important dimension in Swedish inheritance taxation, particularly concerning its changes over time. The corporate tax code had long contained some relief in the valuation of a firm's machinery, inventories, and stocks-in-trade. 12 However, in the 1970s, a series of wide-ranging reliefs for the inheritance of closely held (private) companies were introduced. These reliefs applied to both gifts and bequests irrespective of whether companies were sole proprietorships (enskild firma), partnerships (handelsbolag), or privately held joint-stock companies. That is, these new generous valuation rules applied to the net assets of all firms not listed on the stock exchange.
Specifically, in 1971, a small conditional tax concession of 10 per cent of the inheritance tax on the recipient's lot was introduced. More important, from 1974 onwards, heirs had an option to undervalue stocks-in-trade and inventories. These new valuation rules stipulated that the lowest of either the acquisition cost or replacement value should be used as a basis for taxation. An additional 5 per cent was then deducted for obsolescence, and, finally, the remaining value was written down to 40 per cent. 13 Lastly, in 1978, the valuation relief for private businesses was extended even further: unlisted firms were then valued at 30 per cent of the booked net equity value (assets less liabilities).
The primary political motivation for the valuation reliefs of the 1970s was to facilitate the intergenerational transfer of family firms. The timing was not accidental but directly related to reforms in the early 1970s that prohibited business owners from borrowing from their own firms. 14 These reforms rendered the payment of inheritance tax expensive for cash-constrained heirs who had to either pay themselves dividends taxed at the marginal income tax rate or sell off parts of the firm's assets or shares to finance the tax payment, thus also necessitating that they pay capital gains tax. 15 Tax rate schedules, taxable limits and exemptions, bequest brackets, and the scope of deductions are important components of inheritance taxation. In Sweden, the inheritance tax always depended on consanguinity, that is, the relationship between the deceased and their heirs, with the spouse and children normally paying a lower tax than other relatives or non-relatives. 16 With the 1884 stamp ordinance, 11 The main rule laid down in the 1914 inheritance and gift taxation ordinance was that beneficiary acquisitions were taxed as inheritance; however, before this time, they were partly included. Individual private pension insurance was exempted from taxation. 12 Du Rietz, Johansson, and Stenkula, 'Swedish labor'. 13 Englund, Beskattning av arv och gåva, p. 62. In the tax rate computations below, we have interpreted the deliberate underestimation of stocks-in-trade and inventories from 1974 to 1977 to be an assessment at 40% of equity. 14 SOU 1998:116. 15 The disallowance of borrowing from one's own firm had dramatic effects. Until 1973, firm owners were allowed to borrow from their own firm, even at a zero interest rate. Because the average inflation rate was approximately 7% in the early 1970s, the real rate of interest on such loans was strongly negative. In contrast, if an heir to a family firm needed to withdraw funds from his/her firm in 1974, such funds were first subject to corporate taxation of 54% and then a marginal tax rate of 74%. To obtain one SEK net of tax in order to meet inheritance tax obligations, the firm then needed a before-tax profit of 1/[(1-0.54)(1-0.74)] = 8.36 SEK. 16 Heirs were divided into three (sometimes four) classes. Class I, which had the lowest tax rate, included the surviving spouse, children, and their descendants. Class III comprised juridical persons, such as public utilities  all previous variants of estate taxes, including stamp duties and inheritance lot taxes, were merged into a single tax in the form of a stamp on the total estate value. Initially, the tax rate was basically flat at approximately 0.5 per cent, but in 1895, a progressive tax schedule was introduced.
The Swedish statutory tax rates changed dramatically between 1885 and 2004. Figure 1 depicts the statutory marginal tax rate schedules for immediate family heirs (spouses and children) over the distribution of bequests (expressed as multiples of the average annual incomes of Swedish production workers) in six selected years (where each year is representative of the period in question). Considerable variation exists in both the level and progression in the inheritance tax schedule, with the overall trend being increased levels and progression up to 1991. Tax rates were practically flat and very low until the late 1910s, and were raised somewhat in 1918 and 1920 until the drastic increase in 1933. The largest progressivity appears in the postwar era until the reform in 1992, when the schedule was made basically flat again, but the tax rate was relatively high, and exemptions were limited. and private non-profit foundations and associations, some of which were tax exempt. Class II, strictly speaking, encompassed all other heirs, that is, heirs not belonging to classes I and III. In practice, all other heirs were parents, brothers, sisters, and cousins. Gifts to public authorities and religious communities and foundations promoting research, education, culture, or sports were tax exempt.
The sharp increase in tax rates occurring from 1914 to 1940 is explained by the reforms in 1918 and 1933 (and to a small extent by the 1920 reform). The Social Democrats gained governmental power in 1932. A bill to introduce an estate tax with the inheritance tax was rejected by Parliament, and instead, the existing inheritance and gift taxes were raised. 17 An estate tax was added to the existing inheritance tax in 1948. The two taxes were combined so that the estate was taxed first and the tax payment was then deducted from the estate before the inheritance lots were distributed and taxed. 18 The estate tax was levied on the total net value of the estate after the deduction of certain tax exempt items, such as marital property (half of the estate in the case of a surviving spouse) and a tax-free amount.
In 1959, the estate tax was abolished, and inheritance tax rates were sharply increased. The top tax rate for children and spouses increased to 60 per cent (65 per cent for classes II and IV). The new tax schedules applied until 1970.
The 1970s and 1980s observed further increases in the inheritance tax rates, as also reflected in figure 1. The earlier taxable limits (bottenbelopp) were changed to general deductible exemptions (grundavdrag), and the number of brackets was reduced, which resulted in a small tax increase. 19 These raises in statutory rates were accompanied by alleviations in the valuation of some assets. For example, in 1971, relief in the valuation of private (unlisted) firm assets in the estates was introduced and, from 1978 onwards, the taxable net worth of private firms (assets less liabilities) was further reduced to no more than 30 per cent of the book value of firm equity.
The first reduction in tax rates was enacted in 1987. Specifically, the number of inheritance tax brackets was reduced, and tax rates were adjusted downwards. In 1991, tax bracket boundaries were adjusted upwards in response to the (partly inflation-driven) sharp increase in property values.
In 1992, inheritance tax rates were greatly reduced, and bracket boundaries were adjusted upwards. The tax reduction was motivated by the very high level of inheritance tax in Sweden compared with other countries and the perceived need to reduce the taxation of capital more generally. 20 Figure 1 shows that the top marginal tax rate was reduced from 70 to 30 per cent. The basic exemptions had also been increased several times. The inheritance tax was removed for bequests to spouses in 2003 and fully abolished in 2004. 21 Capital gains taxes also rose if heirs did not possess sufficient cash to pay the inheritance tax and if they sold off assets to finance the tax payment. If the deceased had owned the inherited assets (family firms or other non-corporate assets) for five years or more, capital gains were tax free until 1966. From 1967to 1975, 10 per 17 SOU 1957 A highly progressive income tax schedule was also introduced in 1948 (Du Rietz et al., 'Swedish labor'), and a new wealth tax schedule more than doubled the statutory wealth tax rates (Du Rietz and Henrekson, 'Swedish wealth taxation'). 19 If the inheritance lot was below the taxable limit, no inheritance tax was paid. If the inheritance lot exceeded the taxable limit, the entire lot was taxed. 20 SOU 2002:52, p. 18. 21 The tax was abolished as of 17 Dec. 2004, not 1 Jan. 2005, which was the date originally decided by Parliament. This earlier date was motivated by a concern for the heirs of the Swedish victims of the tsunami disaster in the Indian Ocean on 26 Dec. 2004. More than 500 Swedes, most of them on vacation in Thailand, were killed in the disaster. cent of the capital gains was added to the heirs' personal income tax base and taxed at the marginal income tax rate. Further, the taxable share of long-term capital gains increased to 40 per cent during 1976-90, and, after the tax reform in 1990-1, all capital gains were taxable at a flat rate of approximately 20-30 per cent depending on the type of asset (dwelling or financial asset).

II
How much inheritance tax have heirs in Sweden paid since the end of the nineteenth century to the present? Did tax payments differ across different sizes of bequests and different types of inheritance? This section answers these questions by presenting calculations of the effective average inheritance tax rates for each year during the period 1885-2004.
We calculate tax rates for different bequest sizes in two synthetically constructed estate types: an individual non-family-firm fortune, denoted E, and an entrepreneurial firm inherited by the younger generation in the family, denoted E f . Associated with each of these two estates are the taxable bequests received by the heirs, B and B f . Taxable bequests typically differ from estates owing to the basic exemptions, deductions, and valuation discounts described in the previous section. Taxable bequests also differ from estates because the number of heirs n is usually larger than one, indicating that B = E/n, where E is the estate net of exemptions, deductions, and valuation discounts. Tax rates are computed for the case of two child heirs (n = 2), each inheriting an equal share of the remainder of the estate with no surviving spouse. 22 The effective average inheritance tax rate, Bt , is then defined as the total payment in year t of inheritance, with gift and estate taxes as a percentage of the original estate, that is: For an inherited family-firm estate, the effective average inheritance tax rate becomes: where It denotes year-specific inheritance tax schedules presented in the previous sections. During the period 1948-58, heirs also paid estate taxes, and the estate tax payment Et E was deducted from the taxable bequest. In the case of nonfamily-firm wealth, this deduction implies the following effective inheritance tax: In addition to these taxes, an additional tax related to inheritance taxation was capital gains taxation. Such taxation applied when heirs of family businesses had to sell off part of the company to be able to pay the inheritance tax. 23 Note that if bequests were previously transferred as gifts, the tax was usually not reduced because the basic exemption was lower; thus, the tax rates were identical. Gift taxation could also not easily be reduced by transferring ownership of a company through a combination of inheritance and multiple gifts owing to the summation rules discussed above. However, the inheritance tax was not immediately payable. Rather, it could be paid in instalments over a period of 10 years.
We calculate the effective inheritance tax rates for individual fortunes and family firms for four different sizes of estates: super-large, large, medium, and small. 24 The super-large estate is established at a level corresponding to 1,000 times the average worker's annual salary. In 2004, this level corresponds to an estate worth SEK 266 million (approximately EUR 30 million). The large estate corresponds to 100 times the average worker's salary, which results in a value of SEK 26.6 million in 2004 (approximately EUR 3 million). The medium estate is established at 10 times the average worker's annual salary, which amounted to SEK 2.66 million in 2004 (approximately EUR 300,000). The small estate is a 'middle-class' estate, established at the level of the average taxable wealth in Sweden in 2004 (SEK 622,000 or roughly EUR 70,000), corresponding to 2.5 times an average worker's annual income. We compute the inheritance tax rates only for non-family-firm wealth in this category. Figure 2 presents the evolution of the effective inheritance tax rates, B and B f . A clear inverse U-shaped pattern is evident over the period, and five distinct phases can be distinguished. Effective tax rates were very low before the First World War and then were raised in a second phase that lasted until the early 1930s. The inheritance tax reform of 1933, when statutory tax rates were sharply increased, introduced a third phase. A fourth phase began with the tax reform of 1948, when an estate tax was also introduced and used in combination with the inheritance and gift taxes. In the 1950s and 1960s, the inheritance tax continued to increase, primarily because of an inflation-driven bracket creep in the nominal tax schedule. Effective tax rates then decreased in the 1970s for family firm bequests because of the extensive valuation relief for unlisted corporate assets described in the previous section. Inheritances of non-firm assets remained largely unchanged throughout the 1970s. A fifth phase began in 1992, when the government dramatically reduced 23 In practice, selling off shares or assets may not always have been possible, and heirs had alternative ways to finance their tax payments, for example, paying out extra dividends, taking salaries, or taking loans. In extreme cases, the comprehensive inheritance tax, including the direct inheritance tax plus indirect inheritance taxes in the form of extra income tax and social security fees, could be high enough to exceed the total firm equity. We denote KG as the amount of realized capital gains received on assets held for more than five years (which was typically the case with family firms). KG was taxed as income beginning in 1967 (at variable income tax rates, Yt ). We then add Yt KG to the numerator in equation (2). Before 1966, KG was tax exempt, and between 1967 and 1975, one-tenth of KG was added to labour income and taxed according to labour tax schedules, Yt , typically at a marginal tax rate of approximately 75-85% for high incomes. Between 1976 and 1990, four-tenths of KG was taxable at the labour tax schedule. After the tax reform in 1991, 100% of KG was taxable at a flat tax rate, typically 30%. The effective inheritance tax rate, including capital gains tax, was then Bt = ( It B + Yt KG)/E. 24 Note that these names more closely refer to the relative size of firms rather than individual fortunes; a small family-firm estate corresponds to the net wealth of an individual in the 95th percentile of the Swedish personal wealth distribution. Effective inheritance tax rate (%) 1 8 8 0 1 9 0 0 1 9 2 0 1 9 4 0 1 9 6 0 1 9 8 0 2 0 0 0 1 8 8 0 1 9 0 0 1 9 2 0 1 9 4 0 1 9 6 0 1 9 8 0 2 0 0 0  Figure 2. Effective inheritance tax rates on non-firm and family-firm assets,  Notes and sources: The sizes of the different estates were the following in 2004. A super-large estate was worth SEK 266 million (EUR 30 million), a large estate was worth SEK 26.6 million (EUR 3 million), a medium estate was worth SEK 2.66 million (EUR 0.3 million), and a small estate was worth SEK 0.67 million (EUR 74,000). Note that there is no calculation of inherited family firms equivalent in size to 'Small estates'" or similar. See text for details. Data are from online table S2. the statutory tax rate schedules, which ranged from 10 to 60 per cent (with many brackets), to only three brackets that ranged from 10 to 30 per cent. Irrespective of their size, this policy reduced the effective tax rate to single-digit percentage points for inherited family firms. Tax rates then remained at this low level until the repeal of the inheritance tax in December 2004.
The evidence in figure 2 also clearly shows the distinction in the tax treatment between heirs of family firms (and other private, that is, unlisted, corporate wealth) and heirs of other assets from the 1970s. In 1973, heirs of super-large estates paid an effective inheritance tax of approximately 60 per cent regardless of the type of estate. In the next year, 1974, an heir of a similar non-corporate fortune paid 62 per cent, whereas an heir of an equally valuable family firm paid a mere 24.7 per cent-a horizontal tax wedge of 37 percentage points generated virtually overnight. 25 Although such differential treatment has been well known among tax lawyers, some politicians, and certainly family firm owners in Sweden, it has received little attention in the academic economics literature on inheritance taxation. Indeed, we have found no previous investigations of this horizontal inequity in the inheritance tax or its effect on related phenomena, such as wealth accumulation or tax avoidance activities. Finally, the importance of realized capital gains taxation is reported in online table S2. In general, this additional tax burden does not change the overall pattern of the inheritance tax over time or across different bequest sizes and types. At most, when applied to the heirs of large and super-large estates in the 1970s and 1980s, the capital gains tax raised the effective inheritance tax by one-tenth.

III
What were the main determinants of the dramatic changes in Swedish inheritance taxation during the twentieth century? In particular, what factors explain the sharply increased tax schedules in the interwar and early postwar periods and the equally sharp tax cuts, first, for family firm heirs in the 1970s and then for all other large bequests in the early 1990s? This section presents facts and speculation regarding the main factors that could account for these tax increases. Specifically, we discuss the role of ideology in comparison with more tangible economic or geopolitical factors in shaping inheritance taxation. We also relate the inheritance tax to avoidance behaviour, a topic that has received considerable attention in the public debate regarding inheritance taxation since the nineteenth century. Furthermore, we evaluate whether the alleged motives for high inheritance tax rates produced the desired outcome and why the inheritance tax lost its legitimacy and political support. We noted in the previous section that effective inheritance tax rates were historically low in Sweden until the late 1910s and that they thereafter increased in a stepwise fashion until 1971. Was this trend in inheritance tax policy exceptional or consistent with a common international trend in inheritance taxation? Studying the evolution of inheritance tax rates in 19 countries from the eve of industrialization to 2000, Scheve and Stasavage show that inheritance taxes were invariably very low in all countries before 1900. 26 Notwithstanding large cross-country differences in the top marginal rate, inheritance taxation increased sharply in all 19 countries in the twentieth century. Table 2 shows that Sweden belongs to a group of only six countries in which the top inheritance tax rate exceeded 60 per cent for an extended period (40 years or more). 27 This high rate can be compared with the far lower inheritance tax rates in the seven countries in Scheve and Stasavage's dataset that were attacked or occupied by Nazi Germany or the Soviet Union during the Second World War. 28 In 1950, the average top rate in these seven countries was a mere 20 per cent compared with the average of 75.5 per cent for the six countries in table 2.
Scheve and Stasavage test two alternative explanations for the significant inheritance tax increases occurring in the twentieth century: first, the extension of suffrage and redistributional pressures arising with it and, second, political conditions created by mass mobilization in connection with the two world wars when the broader population demanded increased pecuniary sacrifices by the rich while the people sacrificed their bodies in the wars. Scheve and Stasavage find no evidence of the importance of extended suffrage but strong support for the mass mobilization hypothesis.
Regarding the situation in Sweden, the effect of wartime events and the ensuing large-scale mobilization appears to have influenced the taxation of high wealth and inheritances, but this tendency is not as stark as that found elsewhere. One 26 Scheve and Stasavage, 'Democracy, war, and wealth'. The inheritance tax was low in all countries except the UK and New Zealand, where the top inheritance tax rates in 1900 were 8% and 5%, respectively. 27 The other Anglo-Saxon countries also had comparatively high top inheritance tax rates by the early 1950s: Australia, 28%; Canada, 54%; Ireland, 54%; and New Zealand, 56%. 28 Scheve and Stasavage, 'Democracy, war, and wealth'. possible reason is that Swedish government spending did not increase directly after either world war, which stands in contrast with the upward displacement of government spending in the UK, particularly after the First World War. 29 Another reason concerns the timing of the major hikes in inheritance taxation. The 1918 increase occurred during the First World War; however, Sweden's two largest tax increases, in 1933 and 1948, are not immediately temporally tied to the two world wars. An examination of the views of Swedish politicians during the world wars, as reflected in the writings and debates in contemporary parliamentary print, offers partial support for the link between inheritance taxation and the increased wartime needs and sacrifices of the broader public, as suggested by Scheve and Stasavage. Motivated by the need to finance greater defence spending, the government proposed an increase in inheritance and gift tax rates in 1917. 30 The raise was intended for large inheritances, and the government conjectured that income from inheritance taxation would then increase by 25 per cent. During the Second World War, income and wealth tax progressivity was sharply increased, once again motivated by the need to offset the worsened wartime conditions for the population. However, the inheritance tax remained unchanged and, in fact, it was explicitly excluded from these tax increases because it was unclear whether the state actually needed the additional revenue and it was not deemed possible to arrive at a proper assessment given the exceptional wartime conditions. 31 In other words, mass mobilization appears to be one explanation for the tax increases in Sweden, but it cannot explain the largest tax increases that occurred either between the wars or well after the Second World War. Two additional possible explanations that relate directly to Sweden's economy and political institutions can be put forward. One explanation concerns the role of ideology and the politics of government. The electoral successes of the Social-Democratic Party in the early 1930s also involved a broad public debate in Sweden concerning inequality and the (un)fairness of the wealth distribution and inheritance flows. Leading Social Democrats were well aware that a high inheritance tax could impair the incentives of entrepreneurship and firm formation. However, consistent with Schumpeter's Capitalism, socialism and democracy, many Social-Democratic intellectuals believed that the large industrial corporation was the major unit of production and that an inexorable movement in capitalist societies towards progressively larger companies was occurring. 32 If these beliefs were true, individual entrepreneurship and new firm formation would wane in importance. The leading Social Democrat, Ernst Wigforss, Minister of Finance during 1925-6 and 1932-49, even maintained that, in the long run, large industrial corporations should be converted into 'social enterprises without owners'. In these enterprises, individuals could still be shareholders, but the shareholders were no longer residual claimants. Moreover, in such enterprises, wages should be established in wage negotiations, dividends should be related to the level of interest rates in capital markets, and all excess profits should remain with the companies. 33 29 Durevall and Henrekson, 'Futile quest'; Peacock and Wiseman, Growth of public expenditure. 30 Prop. 1917:167. 31 See SOU 1957:48, pp. 21ff. 32 Schumpeter,Capitalism;Wigforss,Efter välfärdsstaten. 33 See also Johansson and Magnusson,LO andra halvseklet, Digging deeper into the ideological domains of Swedish Social Democracy, Wigforss even authored a critical report on wealth inequality and inheritance taxation in 1928. In his report, Wigforss stated that 'the inheritance tax can be extended and reformed to become a means by which large fortunes are curtailed, at least to the extent that these fortunes do not emanate from the industriousness and thrift of their owners'. 34 Wigforss also asserted that as a means of redistribution, the inheritance tax is more likely to enjoy popular support than (highly) progressive income and wealth taxes. The alleged reason was simple: inherited wealth is not acquired through one's own socially valuable actions and, therefore, it is less legitimate. Wigforss explicitly stated that 'the current wealth distribution cannot be seriously defended, notwithstanding how much one emphasizes the importance of incentives for thrift, diligence and entrepreneurship'. 35 Wigforss further maintained that the British inheritance tax rate of 40 per cent in the 1920s on large inheritances was insufficiently high 'to whittle away the large fortunes and eliminate the cleavage between rich and poor'. 36 Wigforss acknowledged that sharp increases in the inheritance tax would dampen the incentives for saving and entrepreneurship. However, Wigforss identified compensatory mechanisms that could offset these effects. Specifically, the incentives for firms to finance investments through retained earnings could be strengthened. Wigforss wanted to open the possibility of paying an inheritance tax on large fortunes with in-kind assets (stocks, bonds, and real estate). This proposal, Wigforss asserted, would provide an avenue for the increased public ownership of production and collective capital formation that he advocated.
Social-Democratic intellectuals, with Wigforss being the most prominent, had thus argued forcibly in favour of a high inheritance tax throughout the 1920s and early 1930s. They tended to favour a scheme used in Italy and Germany for a few years after the First World War, where the inheritance (or estate) tax was higher if the deceased had inherited his wealth than if he had acquired it in his own lifetime. When the Social Democrats then formed a minority government (based on 41.7 per cent of the votes), they tried to introduce an estate tax that would have raised the combined estate and inheritance tax to British levels. This bill was rejected, and the government had to settle for a less radical-but nonetheless doubling-increase in the inheritance tax rate in 1933. 37 The Social Democrats thus wanted and tried to increase the tax up to British levels in the 1930s, but they were unable to do so owing to insufficient parliamentary support.
After the Second World War the situation was very different. In 1940, the Social Democrats received 54 per cent of the votes, but Sweden had a bi-partisan (multiparty) government throughout the war. Thus, during the war, no room remained for them to exploit their position to pursue any ideological goals (even if they were demanded by the electorate). In 1944 the Social Democrats received 47 per cent of the votes (which gave them exactly 50 per cent of the seats in Parliament) and, 34 Wigforss, Egendomsutjämning och arvsskatt, p. 6. 35 Ibid., p. 9. 36 Ibid., p. 28. 37 Prop. 1933:34. The inheritance tax raise was effective as of 1 July 1933. The proposed top marginal estate tax rate was 21%. The Social-Democratic government attempted once more to introduce a less stringent estate tax in the following year. This time it was renamed a stamp duty of up to 7% of the value of the estate. This bill was also rejected by Parliament.
given that the Communist Party received more than 10 per cent, a solid left-wing majority arose and persisted after the 1948 election, when the Social Democrats continued as a one-party government based on 46 per cent of the votes.
Immediately after the Second World War, a government commission on state taxation set to work and they published their final report in late 1946. 38 In this report, the commission maintained that wealth holders should contribute ex post to make up for the extraordinary costs incurred during the war. The commission discussed various ways that such a tax could be implemented and stressed that no harmful economic effects should arise from the taxation. When motivating the enactment of a separate estate tax in 1947 that became effective in 1948 (before the 1948 election), Wigforss stressed that an estate tax was better than a wealth tax because, he argued, an estate tax would not have detrimental incentive effects (that is, it would neither drain firms of liquidity nor lower incentives for entrepreneurs to accumulate assets). He saw such a tax as a delayed wealth tax.
The estate tax-and the resulting sharp increase in the inheritance tax-was fiercely contested by the opposition and numerous motions in both 1948 and 1949 demanded that the estate tax be abolished and that the amounts already paid be refunded. All the motions were voted down. 39 Effective in 1948, the outcome was a top combined estate and inheritance tax rate of 60 per cent. The top rate was thus raised by a factor of 15-from 4 to 60 per cent-in 30 years.
In sum, the mass mobilization hypothesis of Scheve and Stasavage seems to have bearing on inheritance taxation in Sweden, as evidenced primarily by the 1918 raise but also by the parliamentary discussions during the 1940s. 40 The evolution of inheritance taxation in Sweden through the 1960s, however, shows that mass mobilization is not the full story in this non-belligerent country. The largest tax changes were in fact enacted either between or after the world wars, and contemporary politicians also explicitly stated that the wartime events did not affect inheritance tax policy. Therefore, in this section, we have presented evidence in support of a view that ideological motives for redistribution and long-term vertical equity, deeply rooted in the Social-Democratic Party and temporally associated with changes in its political influence, better explain the dramatic increases in inheritance taxation from the early 1930s to the 1960s in Sweden than mass mobilization. Notably, this ideology is also closely aligned with the notion of maximizing the welfare of groups at the bottom of the distribution, at least to the extent that public sentiment towards high inheritance inequalities is sufficiently negative while the impact of the tax on wealth accumulation is sufficiently small. 41 Tax avoidance and the effectiveness of the inheritance tax is another potentially important factor to address in an examination of its evolution over time. Figure 3 displays the evolution of revenue from the inheritance, gift, and estate taxes, that is, B B, divided by either total tax revenue or national income. The share of inheritance tax in total tax revenue hovered at about 2 per cent in the interwar era and 1 per cent in the 1940s and then decreased steadily until the 1990s to a 38 SOU 1946:79. 39 SOU 1957 Scheve and Stasavage,'Democracy,war,and wealth'. 41 Piketty and Saez, 'Theory'. Inheritance taxes as % of total taxes (%) 1 8 9 0 1 9 1 0 1 9 3 0 1 9 5 0 1 9 7 0 1 9 9 0 Inheritance tax revenue as % of total tax revenue (left scale) Inheritance tax revenue as % of national income (right scale)

Figure 3. Revenues of Swedish inheritance and gift taxes, 1885-2004
Notes: Estate tax payments between 1948 and 1959 are classified as inheritance tax payments. Because of lags in estate inventories and tax payments, the taxes still generated revenue for several more years even though tax liability ceased for deaths after 16 Dec.

2004.
Sources: Data on inheritance and gift taxes are from Ohlsson, 'Legacy'. Data on total tax revenue are from Gårestad, 'Industrialisering och beskattning'; and Rodriguez, Offentlig inkomstexpansion. See online table S1.
level of approximately 0.2 per cent-a reduction of over 90 per cent. This level is low not only in absolute terms but also in comparison with international levels: by the end of the 1990s, the inheritance tax as a share of total tax revenue was 0.7 per cent in Germany, 1.4 per cent in the US, and 1.8 per cent in France. 42 In Sweden the share of inheritance tax revenue in national income was much lower, approximately one-fifth of 1 per cent in the interwar and early postwar decades. Subsequently, the inheritance tax revenue fell by approximately half by the end of the century, that is, a decrease of 50 per cent, and the drop was not as severe as the decrease in the share of inheritance tax revenue in total tax revenue. Thus, although the overall fiscal significance of inheritance tax revenue was never large in Sweden, it fell rapidly during the postwar period.
Regarding the evolution of the actual tax base, a recent investigation by Ohlsson, Roine, and Waldenström shows that the amount of inheritances (and gifts) flowing from the deceased to the living in Sweden was approximately 12 per cent of national income just before the First World War. 43 This flow then fell steadily during the interwar period to a level below 5 per cent in approximately 1950, nearly the same level as that in France for this period. 44 Generally, the tax revenue data and estimates of inheritance flows do not suggest that inheritance tax revenue was ever considered an important source of government revenue per se. This finding is consistent with Beckert's findings regarding the historical evolution of inheritance taxation in France, Germany, and the US. 45 42 Beckert, Inherited wealth, p. 273. 43 Ohlsson et al.,'Inherited wealth'. 44 Piketty, 'On the long-run evolution'. 45 Beckert, Inherited wealth. Tax avoidance or tax evasion represents a distortion that should engender a lower tax rate according to standard optimal tax models. A few studies on Sweden document the behavioural responses to inheritance and gift taxes. Eliason and Ohlsson find that heirs seem willing to misreport the dates of death of their parents to reduce expected inheritance tax payments. 46 Ohlsson finds a strong expansion of gifts recorded just before the introduction of the estate tax in Sweden in 1948, a pattern that clearly reflects avoidance behaviour. 47 The trends in inheritance tax avoidance in Sweden have, to our knowledge, not been studied previously and, in this article, we can only speculate about these trends on the basis of both qualitative and partly anecdotal evidence and a quantitative exercise that relates observed tax revenue to newly estimated inheritance flows.
The qualitative evidence suggests that both avoidance opportunities and actual avoidance behaviour increased along with the increase in effective inheritance tax rates in the 1930s and 1940s. Policy makers at the time seem to have noticed this pattern, particularly during the estate tax regime in 1948-58, when it soon became clear that the estate tax did not raise as much revenue as had been originally estimated. 48 The study by Ohlsson shows how the amount of gifts exploded just before the estate tax was introduced in 1948. 49 In the 1950s, the establishment of tax-exempt family foundations, holding companies, and limited partnerships became another well-known measure to avoid the estate tax. 50 Life insurance policies with a designated beneficiary constituted another response to minimize inheritance tax receipts. Disbursements from life insurance were exempted from inheritance (and estate) taxation and, if the insurance premium was not deducted from current income, disbursements were tax exempt for the beneficiary. 51 Debt expansion was also used by wealthy business owners to avoid taxes, as any debt was fully deductible from the (tax-assessed) value of the assets. In particular, diversification into real estate offered a typical way to reduce the inheritance tax. 52 After the deregulation of credit markets in the 1980s, opportunities for the elimination of the inheritance tax through the acquisition of highly leveraged assets with a reduced tax-assessed value increased greatly. Although we have not managed to find systematic evidence regarding the use of this and similar avoidance strategies, numerous reports in the business press, particularly during the 1970s and 1980s, note the use of such strategies.
An extreme way to avoid paying taxes was to move either one's wealth or oneself out of the country. Numerous examples show how the wealthiest and most 46 Eliason and Ohlsson, 'Timing of death'. 47 Ohlsson, 'Legacy'. The international literature on inheritance tax planning and avoidance is also small; see Kopczuk, 'Taxation of intergenerational transfers', for a recent overview. 48 SOU 1957:48, p. 10. 49 Ohlsson, 'Legacy'. 50 SOU 1957:48, p. 10. Feldt, Den blyge entreprenören, documents in some detail the drastic plans considered and measures eventually taken by the Johnson dynasty to avoid being severely affected by the combined effect of the estate and inheritance tax in case of the passing of Axel Ax:son Johnson (1876-1958), the patriarch and sole owner of the industry group. 51 Johansson, 'Arvs-och förmögenhetsskatt', and Johansson and Werding, 'Rate of return', calculate the required rate of return under conservative assumptions for business owners. The required rate of return became unrealistically high relative to industry returns at the time (typically 20% in 1970 compared with average actual returns of 6%). 52 Additionally, real estate was a frequently used way to reduce the wealth tax, which was arguably even more important; Du Rietz and Henrekson, 'Swedish wealth taxation'. successful entrepreneurs and business owners chose to emigrate. 53 In a study of all permissions granted by the Riksbank to transfer assets abroad in connection with emigration during 1965-84, Lindkvist documents 30,000 permissions to emigrants during this time. 54 From interviews with people emigrating during the period 1975-84, Lindkvist also finds that Swedish taxation, including inheritance taxation, was a major consideration in the emigration decision for most respondents. When Sweden lifted all foreign exchange controls in 1989, effectively allowing anyone to transfer assets to tax havens either by using illicit means or by taking up residence in another country, a notable increase in tax-driven capital flight appears to have occurred. 55 Quantitatively assessing the trends in inheritance tax avoidance or evasion in Sweden is difficult. One channel of potential information is offered by relating actual taxes paid with two variants of the underlying tax base: a 'fiscal inheritance flow', reported to tax authorities on inheritance tax returns, and an 'economic inheritance flow', a theoretically computed inheritance flow based on data on the aggregate stock of private wealth, adult mortality, and the life-cycle profile of wealth. 56 The left panel of figure 4 shows the ratio of inheritance tax revenue to each of these tax bases. During the entire period, the 'fiscal flow' ratio, that is, the revenue to taxed inheritances, exceeds the 'economic flow' ratio, that is, the revenue to macro-estimated inheritances.
Does the difference between the 'fiscal flow' and 'economic flow' tax rates reflect tax avoidance? We cannot be certain about such a conclusion because other data differences exist between the two sources, for example, asset composition and valuation procedures. 57 However, the incentives to avoid taxation should increase as the statutory tax rate increases. The right panel of figure 4 shows that the difference in effective tax rates between large and small bequests strongly correlates with the difference between the 'fiscal flow' and 'economic flow' tax rates. When the incentive to avoid inheritance taxation was high for individuals inheriting a significant amount (that is, a large difference between the tax on large and small bequests), we also observe a substantial difference between the amounts of reported inheritances and the amounts estimated based on other statistical and demographic data. This result suggests that avoidance behaviour may well have increased substantially in the postwar era and remained high until the repeal of the inheritance tax in the early 2000s. 53 Ruben Rausing, the founder of Tetra Pak, emigrated in 1969, and his two sons followed in 1982. Ingvar Kamprad (the founder of Ikea) and Erling Persson (the founder of H&M) emigrated in 1973 and 1982, respectively. A more recent example is Fredrik Lundberg, who moved to Switzerland in 1985 and lived there until 1993. 54 Lindkvist, 'Kapitalemigration'. 55 Roine and Waldenström, 'Long-run trends', attempt to estimate the size of tax-driven capital flight by accumulating statistical errors and omissions in the balance-of-payments statistics. They find that these errors regularly indicated increased outflows during the 1990s and 2000s, plateauing at a number near one-fifth of one year's national income. 56 Data on these two flows come from Ohlsson et al., 'Inherited wealth'. See also online tables S1 and S3, where all tax bases and tax rates are reported. Note that we observe only 'fiscal flow' inheritance in several years in relation to surveys by the authorities of the inheritance tax system. 57 Ohlsson et al., 'Inherited wealth', note that the tax-reported inheritance flow misses several important components, including the unpaid collective insurance savings of the deceased, such as occupational pension funds. The legitimacy of taxes among taxpayers is sometimes at odds with the optimal tax profile. Stable tax bases may be economically efficient, but the possibility of taxing them also requires political support and a general acceptance of the tax. Exactly what determines political support for a tax varies from case to case, but horizontal and vertical fairness perceptions and tax visibility can be important. 58 In this section, we examine the dimensions of the legitimacy of the Swedish inheritance tax as an additional factor in understanding its change over time and, in particular, its demise at the end of the twentieth century. Figure 5 shows two important trends. The solid line shows the basic exemption level for tax eligibility as a share of a worker's annual salary, which indicates how much of bequests were affected by the tax in real economic terms. Although the exemption level remained low throughout the period, a secular decline occurred with the lowest levels reached from the 1970s onwards. The dashed line in the figure, in contrast, shows the wealth level required to be subject to the top rate as a multiple of a worker's annual salary. This level is recurrently very high when tax rates were sharply increased, but, in this case, there is also a secular decline, largely as a result of bracket creep. In 1992, a sharp break occurred when the tax policy was reformed to apply the highest inheritance tax rate to modest-sized fortunes as well. In the last year of the tax, the exemption level was merely one-quarter of an annual production worker's income (SEK 70,000), and the top marginal rate was reached at an inherited amount of just over twice the annual income of a production worker. The basic exemption level of SEK 70,000 in Sweden in 2004 can be compared with the basic exemptions in other countries. In the US, the basic exemption level was 150 times larger, at USD 1.5 million (SEK 10.5 million), with only the top two percentiles in the estate distribution paying any tax compared with approximately one-quarter of all heirs in Sweden. 59 In France, the basic exemption level was EUR 300,000 (SEK 2.7 million), and in the UK, GBP 263,000 (SEK 3.4 million), with both countries having exemption levels more than one order of magnitude greater than Sweden (38 and 48 times greater, respectively). Figure 6 shows that fairly modest inheritances were also hit by a high wealth tax rate. As shown, the threshold for paying a 30 per cent inheritance tax at the margin more or less continuously declined from 1959 until the repeal of the tax. For the last three decades of the tax, the threshold was consistently below four times and occasionally even below two times an average production worker's annual income. Thus, over time, exemptions were reduced, and the inheritances that were affected by high effective inheritance taxation became small. Consequently, members of the broad middle class, who by no means considered themselves wealthy, were increasingly eligible to pay inheritance tax.
Safety valves were increasingly built into the system from the 1970s onwards, with the intention of mitigating, or even completely eliminating, the inheritance tax on large fortunes. As shown in figure 2, in the 1970s, heirs of unlisted business equity were granted a series of reliefs that substantially reduced their effective inheritance 59 SOU 2004:66, tab. 2.3. Lower bequest limit at which inheritance tax is 30 % / APW 1 9 5 0 1 9 6 0 1 9 7 0 1 9 8 0 1 9 9 0 2 0 0 0 In a theoretical model (loosely tested against Swedish and Danish data), Roine suggests that such tax policy results from a political equilibrium consisting of a coalition between the poor and the super-wealthy, where the latter are granted tax avoidance possibilities and, in return, favour the increased taxation of people of more modest wealth and income. 60 This model is suggestive in that it may explain why a norm was created that made politicians adhere to an inheritance tax system that became increasingly onerous for the middle class. The system was not reformed until the tax had lost its legitimacy altogether.
A major motivation for the substantial reliefs in the valuation of assets in private firms was the situation created by the prohibition of borrowing from one's own firm to pay the tax: not only did the tax become extremely cumbersome (sometimes 10 SEK or more of before-tax profit was required to pay one SEK of inheritance tax), but selling the firm instead became extremely advantageous. Capital gains taxation was still favourable (zero until 1966 and a maximum of 8 per cent of the sales value until 1976), but the sale of the firm had to be effected before the death of the owner-entrepreneur. Thus many entrepreneurs sold their firms pre-emptively to avoid inheritance taxation and the firms tended to be sold to tax-favoured investment companies. 61 Moreover, in many cases, the former owners emigrated to a low-tax country. This situation was considered a problem, motivating the reliefs instituted in 1974 and in 1978. 62 In addition, politicians clearly had a very different view of the inheritance tax levied on traded assets. In this regard, the Social-Democratic Minister of Finance, Gunnar Sträng, clearly stated that these firms would not be harmed if the heirs had to divest to pay their inheritance tax. 63 Naturally, this policy was even more applicable to other financial assets, such as bank deposits and bonds, and to real estate.
Another example of increased opportunities for the super-wealthy to avoid inheritance taxation was the deregulation of credit markets, the removal of foreign exchange controls in 1989, and EU membership in the mid-1990s.
In addition to these trends, the 1980s and 1990s in Sweden also witnessed a number of publicized incidents that exposed problems with the inheritance tax, typically showing how heirs had to pay tax at rates that were considerably higher than the relevant statutory rates. Arguably, the most publicized case was the estate of Sally Kistner, the wealthy widow of the founder of the pharmaceutical company Astra and the company's largest individual shareholder. As the Astra stock price fell just before the estate was transferred, the value of the assets fell short of the tax obligations, leading to tax rates above 100 per cent. Consequently, the estate filed for bankruptcy. 64 When Fredrik Lundberg-currently one of Sweden's wealthiest and most powerful industrial owners-migrated to Switzerland in 1985, he explicitly referred to the Kistner case as a motivation for his move. 65 A population survey of attitudes towards taxes in Sweden conducted in 2004 shows that close to two-thirds of the respondents, including a majority of leftleaning persons, wanted inheritance and gift taxes to be either reduced or removed altogether. 66 Popular support for the inheritance tax was thus weak at this point, possibly because of the developments just described. Indeed, in the government bill to repeal the tax, the first two arguments mentioned were the increased problems for heirs in paying rising inheritance tax owing to increased property values and the increased incidence of stock owners delisting their shares to avoid the tax. 67 How can Sweden's repeal of the inheritance and gift taxes in 2004 be explained? The trend discussed above, of an increasing inheritance tax burden on heirs receiving low or moderate bequests and increasing opportunities to avoid the tax among the very wealthiest, generated an overshooting in taxation and reduced the legitimacy of the inheritance tax. The precise timing of the repeal may be related to two specific events. First, deregulation had increased the price of real estate and engendered a booming housing and stock market. 68 As a result, a growing percentage of middle-class heirs had to pay inheritance tax. For example, in the mid-1980s, the marginal inheritance tax rate was 18 per cent on an inheritance equal to the annual income of an average worker. Second, key legislative changes in the late 1990s combined with increasingly innovative strategies further enabled wealthy heirs to avoid the inheritance tax.
The most important feature of the tax system that could be exploited to reduce or eliminate the inheritance (and gift) tax was the fact that debt was valued at 100 per cent, while not only family firms, but also several other assets, notably condominiums and shares listed on the unofficial O-list, had an assessed value far below their market value: 69 30 per cent for the O-listed stock and even less for condominiums. Following the deregulation of credit markets in the 1980s people could start exploiting this asymmetry. Wealth could readily be transferred by selling such assets to descendants at the assessed value rather than the market value, even accepting payment in the form of a promissory note at a rate of interest of less than a third of the market rate. Second, following the abolition of exchange controls in 1989-93, 70 it became much easier to avoid taxation on assets owned in Sweden or received from somebody in Sweden by emigrating. Most importantly, the bilateral tax treaties between Sweden and the UK and Switzerland, respectively, entailed that no tax was levied on inheritances and gifts received from Sweden by Swedish immigrants to those two countries.
Thus, although family-firm owners were favourably treated from the 1970s, new laws combined with more efficient avoidance strategies during the 1990s enabled people with the highest wealth to avoid paying inheritance tax. In other words, while broad layers of the population over a short period in the mid-to-late 1990s experienced rapid increases in their inheritance tax burden, the heirs of very large fortunes experienced the opposite development.
The government did not attempt to argue or explain to the electorate that substantially reducing the inheritance tax rate and applying it to a broader asset base with larger exemptions would be favourable. Although the distortionary effects resulting from evasive behaviour may have been growing, the system remained unreformed. Ultimately, the inheritance tax lost its legitimacy among people because it became regarded as a voluntary tax for the very wealthy, 71 while simultaneously hitting a large share of middle-class heirs (through low basic exemption levels), who also soon reached the top marginal tax rate (recall figure 1).
The abolition of the inheritance tax by a Social-Democratic government may seem paradoxical. In addition to the above factors, an accidental element may have been involved. In an interview conducted by tax law professor Sven-Olof Lodin, then Prime Minister Göran Persson (1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006) acknowledged that the decision to repeal the tax was part of a horse-trading scheme between the Social Democrats and the Left Party. 72 The government had concluded that reform of the wealth tax and/or the inheritance tax was necessary. The wealth tax was generally deemed to be more distortive, but polls showed that the inheritance tax was more unpopular. Göran 69 For this reason, no new firms that went public chose the official so-called A-list when they went public, and soon many of the largest and best-known firms began to move from the A-to the O-list. H&M, the most highly valued firm on the Stockholm Stock Exchange, was never on the A-list and, just before the inheritance tax was abolished, a mere 10% of all firms on the Stockholm Stock Exchange were on the A-list. 70  Persson had decided that one of the tax policies had to be repealed. In the negotiations between Persson and Lars Ohly, the then leader of the Left Party, Ohly was given a choice between the abolition of the inheritance tax and the abolition of the wealth tax, and Ohly chose the inheritance tax. 73 IV This article presents new facts regarding the long-run evolution of inheritance, gift, and estate taxation in Sweden from 1885 to the repeal of the taxes in 2004. Our contribution is twofold. First, we compute the annual effective average inheritance tax rates for different bequest sizes and for two different types of inherited assets: personal net wealth and family-firm equity. These series account for all relevant institutional tax rules and stand in complete contrast to the most commonly used metric of inheritance taxation in previous studies: the top marginal tax rate. In fact, disregarding the national institutional specificities in inheritance, taxation can have significant consequences for the assessment of inheritance tax levels and trends. The long-run evolution of effective inheritance tax rates exhibits an inverse-U shape over five broad phases. Following an initial low-tax phase beginning in the 1880s, a second phase began towards the end of the First World War and lasted until 1933. The third phase emerged as the newly appointed Social-Democratic government doubled the tax rate for large bequests. A long fourth phase then began in 1948 with the introduction of an estate tax in addition to the inheritance tax and, when the estate tax was abolished a decade later, inheritance tax rates were raised commensurately. At this time, Swedish inheritance taxation reached an international high, with top marginal tax rates of 70 per cent and average effective tax rates between 50 and 60 per cent for large estates (with non-deductible latent capital gains taxes). Tax levels remained high until the early 1990s, with the exception of the drastic tax cuts in the 1970s for inherited family firms. The fifth and final phase of the Swedish inheritance tax began with the overhaul of the tax policy in 1992, when the taxation of large inheritances was reduced from 60 to 30 per cent (although without changing-and even slightly raising-the tax on small, middleclass bequests). The final phase lasted until the tax was repealed in 2004.
Our second contribution is our attempt to explain the changes in inheritance taxation by drawing on a combination of quantitative and qualitative evidence. We propose four major factors that appear to have played a role in this process. First, we trace the expansion of inheritance taxation to domestic politics and the ideological momentum of the Social Democrats governing the country continuously from the early 1930s until the mid-1970s. Inheritance and gift taxes were not particularly important as a source of revenue for the government. With few exceptions, less than 2 per cent of Sweden's total tax revenue was generated through such taxation, and in the last 40 years before the abolition of the tax, the share was approximately onetenth of this level. Inheritance and gift taxation was instead primarily motivated by distributional concerns, based on an urge to balance large inequalities in opportunity arising from inherited wealth at the top of the wealth distribution. 73 Particularly towards the end of his time in office, Göran Persson dominated the government and, on crucial issues, he sometimes acted largely independently. This course of events has also been confirmed by Kjell-Olof Feldt, Social-Democratic Minister of Finance in 1982-90 (email correspondence, 19 April 2015.
The electorate apparently accepted an excess tax burden as the price required for a more even distribution of wealth, a policy that thus could be aligned with maximizing the welfare of the least well-off groups in society, at least in the early era. Still, the evidence on increasing tax avoidance and falling revenues during the postwar period, despite sharply increased tax rates and a reduced exemption level, casts doubt on the effectiveness of the inheritance and gift tax from the late 1940s until the 2004 repeal.
Second, mass mobilization related to the world wars appears to have influenced inheritance taxation in Sweden. The relevance of this factor even in a neutral country in both world wars lends support to the mobilization effects on taxation of the rich suggested by Scheve and Stasavage. 74 However, this effect is not as strong in Sweden as in the belligerent western countries, and it cannot explain the most substantial changes in Swedish inheritance taxation.
Third, tax avoidance, and probably tax evasion, appear to have increased in scope over time and in the level of the inheritance tax. This finding indicates that the economic distortion generated by the tax grew over time and that such distortion was possibly at its largest towards the end of the past century, immediately before the tax was ultimately repealed.
Finally, in contrast to the more exclusive regimes in, for example, France, the UK, and the US, the inheritance tax regime in Sweden affected a relatively large portion of the population. Although the top tax rates for heirs to family firms and other large fortunes were reduced in the final decades of the twentieth century, the effective taxation of middle-class bequests even increased somewhat. The combination of broad taxpayer coverage and rapidly growing opportunities for the rich to reduce their effective tax undermined the legitimacy of the inheritance tax. This loss of legitimacy was also observed in tax attitude surveys. Ultimately, this loss of legitimacy explains why there was so little support for the tax and so few objections when it was finally abolished in 2004 by a Social-Democratic minority government backed by the Left Party.