PART IV - TRENDS IN GLOBAL WEALTH INEQUALITY

Global wealth inequality estimates are scarcer than for global income inequality and subject to caution

The available data on wealth inequality is much sparser than for income inequality, especially at the global level. It is therefore more difficult to provide a complete picture of how global wealth inequality has evolved over the past few decades.

We want to be very clear about this: available data sources make it impossible at this stage to properly estimate the level and evolution of the global distribution of wealth. We can to some extent estimate the global distribution of income and its evolution, as we have tried to cautiously show in Part II of this report. The situation is different for wealth. As we have shown in Part III of this report, there are very large areas of the world—particularly in Africa, Latin America, and Asia—where we are not even able to properly measure the aggregate level of national wealth and its decomposition into private and public property, foreign wealth, and natural capital. We first need to make more progress on the measurement of total wealth and its changing structure before we can construct estimates of distribution of private wealth among individuals.

A number of magazines (most notably, Forbes) do publish global rankings of billionaires, and some financial institutions (for instance, Credit Suisse) have combined billionaire data with other data sources to estimate global distributions of wealth. Typically these studies find that top wealth holders have been rising at very high speed in recent decades—substantially faster than the size of the world economy—and below we will agree with this general conclusion. However the methodologies used by Forbes and by these institutions often lack transparency; in particular, they do not release their raw data sources and detailed computer codes. It is impossible therefore to reconstruct their statistical results. This is not merely a technical question; methodological choices can indeed have a large impact on the measured evolution of wealth inequality, and transparency of methods and sources is critical if we want to reach some agreement about inequality facts.

In the context of the WID.world project, we choose to proceed in a gradual manner and to release wealth inequality series solely for the countries for which raw sources allow us to do so in a satisfactory manner. Ideally, one needs to combine household wealth surveys together with wealth rankings and administrative fiscal data (coming from both the income tax, using the capitalization method, and the inheritance tax, using the estate multiplier method) to be able to properly estimate the distribution of wealth and to confront sources in a transparent way. At this stage, these conditions are satisfied only for a handful of countries—most notably, the United States, a number of countries in Europe (in particular, France, the UK, and Spain), and to a lesser extent China (where we have access to household wealth surveys and wealth rankings, but where access to fiscal data is extremely limited). We have also produced estimates of wealth inequality for Russia and the Middle East, but they are more fragile, and we do not use them to produce global wealth estimates in this report.

Our global wealth inequality estimates since 1980 therefore combine data from three large regions: the United States, China, and Europe. Europe itself is represented by three countries (France, Spain, and the United Kingdom), which on the basis of other countries for which we have wealth inequality data (in particular, Sweden and Germany) appear to be broadly representative. Starting from 1987, we can also compare our results with the Forbes billionaire rankings, which provide a better coverage of countries, though only for a tiny, extremely wealthy part of the population, and with little knowledge of how this information was collected.

Available data show that global wealth inequality is extreme and on the rise

At the global level (represented by China, Europe, and the United States), wealth is substantially more concentrated than income: the top 10% owns more than 70% of the total wealth.¹ The top 1% wealthiest individuals alone own 33% of total wealth in 2017. This figure is up from 28% in 1980. The bottom 50% of the population, on the other hand, owns almost no wealth over the entire period (less than 2%). Focusing on a somewhat larger group, we see that the bottom 75% saw its share oscillate around 10%. Wealth concentration levels would probably be even higher if Latin America, Africa, and the rest of Asia were included in the analysis, as most people in these regions would be in the poorer parts of the distribution. We leave this to future editions of the World Inequality Report. (Figure 4.1.1)

We compare in Table 4.1.1 the growth rates of the different wealth groups between 1980 and 2017 (all growth rates are expressed in real terms—that is, after deduction of inflation). A number of striking findings emerge. First, one can see that average wealth has grown faster since the 1980s than average income, reflecting the general tendency of wealth/income ratios to rise in most countries, as documented in Part II of this report. Between 1987 and 2017, per-adult average income has increased at 1.3% per year at the world level, while per-adult wealth has increased at 1.9% per year.

Next, if we now look at the top of world wealth distribution—as measured by the Forbes billionaire rankings—we find that the top wealth holders’ share has increased a lot faster than average wealth holders: 5.3% since 1987 for the top 1/20 million, and 6.4% for the top 1/100 million (see Table 4.1.1). By definition, this is an evolution that cannot continue forever: if top wealth holders were to grow on a permanent basis at a speed that is three to four times faster than average wealth in the world, then billionaires would ultimately come to own 100% of the world’s wealth.

The problem with this billionaire data is twofold: first, as was noted above, it is not entirely clear how it was estimated; next, and most importantly, it is not clear at all whether this pattern of very fast growth holds only for billionaires, or whether it can be extended to multimillionaires. This is crucial because there are many more individuals who own $5 million, $20 million, or $100 million than there are billionaires, and the former command a potentially much larger fraction of world wealth than the latter.

We unfortunately do not know the full answer to this question, but at least our estimates for the US, Europe, and China distribution of wealth provide some interesting insights. We find that the top 1% average wealth in the US, Europe, and China has risen at 3.5% per year between 1987 and 2017 (versus 2.8% for per-adult average wealth and 1.9% for average income). The higher we go in the distribution, the faster the growth: the top 0.1% average wealth has increased by 4.4% per year, and the top 0.01% average wealth has increased by 5.6% per year.

These findings, which were obtained by combining a number of independent data sources (household wealth surveys, income tax data using the income capitalization method, and inheritance tax data using the estate multiplier method, when available), appear to be consistent with the Forbes billionaire data. But they also suggest that one needs to go really very high in the distribution of wealth to see growth rates on the order of 5%–6% per year. If one considers only the top 1% wealth holders as a whole (that is, all individuals with net wealth higher than about €1.1 million in China, Europe, and the United States in 2016), then the growth rate between 1987 and 2017 has been 3.5% per year. This is faster than average wealth growth (2.8% per year), but the gap is not as huge as for billionaires. This suggests at current speed that rising inequality and the divergence of the wealth distribution will take a couple of decades before it takes really extreme proportions. (See below for a discussion of future prospects.) That being said, the direction in which the distribution is going definitely suggests rising concentration of wealth, and there is no evidence that the financial crisis of 2008 had any impact—other than temporary—on this long-run structural trend.

Our results also show that a large share of the growth of global wealth accrued to the top 1% and even narrower wealth groups. As Table 4.1.2 shows, the top 1% captured 37% of per capita wealth growth, more than half of which went to the top 0.1%.

All of this implies growing inequality at the top end of the distribution. Note that the bottom of the distribution has also experienced a significant increase of its wealth, driven by rapid growth in China, as shown by Figure 4.1.2. This pattern is reminiscent of the “elephant curve” of global income growth, showing that the global wealth distribution seems to have evolved in ways qualitatively similar to income. The bottom three-quarters of the distribution saw its wealth increase by a sizeable amount, though less than the world’s billionaires according to Forbes. Between those two groups, wealth growth was at its lowest for the middle class in developed countries. The trends in the wealth growth of different groups have been fairly stable over the last three decades, with narrower wealth top groups experiencing higher growth.

Under a business-as-usual scenario, the top 1% wealth share will increase at 1 percentage point every five years

What will happen to the global distribution of wealth if these trends were to continue for the next few decades? Figure 4.1.3 seeks to answer that question. The top 0.1% wealth owners would progressively catch up with the global wealth middle class, which we define as wealth holders below the top 10% and above the median—that is, 40% of the world population. In 2050, both groups would own the same share of global wealth—that is, 25%. The global wealth middle class comprises 40% of the world population meaning that the top 0.1% wealthiest would be on average four hundred times wealthier than the global middle class. This evolution would take a couple of decades.

The top 1/20 million and 1/100 million of individuals, which comprise about 250 and 50 adults, could respectively own 1.5% and 0.75% of total wealth as soon as 2030, up from 0.5% and 0.25% in the early 1990s. The share of the top 1% would keep on increasing by one percentage point every five years. The shares of the top 0.1% and 0.01% would also grow by one percentage point every five years, meaning that the increase in wealth inequality is in fact driven by these small groups. These groups are much broader than billionaires, but nevertheless quite narrow. (To belong to the top 0.1% or top 0.01% of Europe, the United States, and China in 2016, one needs to own more than €5.2 million or €25.8 million, respectively.)

Global wealth inequality is driven by a large number of forces

As discussed in Part II, global income dynamics are driven by both between- and within-country forces. The rise of private wealth has been faster in large emerging economies than in rich countries, a trend driven by high economic growth and large-scale privatization in transition economies. This tends to reduce global wealth inequality. This effect was more than offset at the top, however, by the rise in wealth inequality within countries. Rising wealth inequality within countries is itself due to a number of factors, including rising income inequality amplified by inequality of savings rates and of rates of return. Other factors, such as the progressivity of taxation, can in turn mitigate or worsen these dynamics. Hence, future global wealth inequality will depend on both catchup growth in emerging economies, and within-country determinants of inequality. We study them at the country level as further described in the next chapters.

We should stress at the onset that there was nothing inevitable about the fact that the very top of the global wealth distribution would rise so much faster than average world wealth beginning in the 1980s. One of the global factors that might have played a role is the larger transfer from public to private wealth that took place in many countries. (See Part II.) To the extent that privatization disproportionately benefited small groups of the population—for example, Russian oligarchs—this can help explain why top wealth holders’ shares rose so fast. It is difficult, however, with the data at our disposal to estimate the global impact of this factor. In particular, there are also some cases where privatization has benefitted mostly the middle class (for example for housing, as we discuss below for the case of the UK, France and Spain). Whether this channel is likely to be important for the future (one might be tempted to conclude that large privatization waves are now behind us) is another important and uncertain issue.

Another potentially important global factors behind booming top wealth is the fact that financial deregulation and innovation might have increased the inequality in rates of return that are accessible to different sizes of financial portfolio. Some of the most convincing evidence for this channel comes from the observed real rates of return on university endowments, which varied from 4–5% per year for the smallest endowments to as much as 8–10% per year for largest ones (after deduction of inflation and management costs) in the United States between 1980 and 2010.²

Again one might wonder whether this corresponds to a specific financial period or whether this will continue in the future (available data suggests that large endowments were still getting very good returns in recent years). Also the governance of personal family wealth involves many other issues than that of large academic capital endowments, so one cannot directly apply these findings. Unfortunately there is too little data available to make similar computations for the highest family wealth.

As we shall see below, however, our country studies do show that differential rates of return—together with differential saving rates—can potentially be an important driving force behind rising wealth concentration. (Box 4.1.1.)

Wealth inequality in the United States has risen rapidly and consistently since the mid-1980s

To fix notions of wealth inequality in the United States, it is perhaps best to first consider the distribution of the country’s wealth in 2012 that is outlined in Table 4.3.1. The average net wealth per family was over $384 000, but this average masks a large heterogeneity. The bottom 90%—a group of almost 145 million families who possess approximately $94 000 on average—collectively own about as much of the total household wealth (22%) as the 161 000 families who are included in the top 0.1%; their average wealth was approximately $82 million, 845 times larger than the bottom 90%. Wealth is much more concentrated than income in the United States, as the top 0.1% wealth share is about as large as the income share of the top 1%.

Rising wealth inequality since the 1980s is almost entirely due to the top 0.1%

Wealth is becoming significantly more concentrated in the United States, but this trend is not the result of tens of millions of Americans seeing a rise in their fortunes. It is rather the spectacular dynamics of a tiny group of the population owning more than $4.4 million—the entry price of the top 1%.

Top wealth shares have risen sharply since the mid-1980s. Indeed, the share of wealth held by the top 10% in 1985 was approximately 63%, the lowest value it had reached since 1917. But by 2012, the wealth share of the top 10% had reached over 77%, an additional 13 percentage points. More than three quarters of all wealth in America was owned by just ten percent of its population.

However, since the mid-1980s, the wealth share of families belonging to the top 10% but not to the top 1% has decreased. In fact, the share of total wealth owned by the top 1% increased at a faster pace (up by around 17 percentage points) than the top 10% between 1986 and 2012 (see Figure 4.3.1a). The rise in the wealth share of the top 1% itself owes almost all of its increase to the growth of the top 0.1% share which rose from 7% to 22% (15 percentage points). The wealth share of the top 0.1% was thus larger than the share of the top 1–0.1% (that is the top 1% minus the top 0.1%) in 2012, having tripled since 1978. Almost all of the top 1% and top 10% increase over the past four decades has been due to the top 0.1% alone.

The recent rises in wealth concentration contrasts with continual reductions over the previous half-century

The significant increase in the wealth shares of America’s wealthiest since the mid-1980s is in direct contrast to the trend that followed the Great Depression. The Roaring Twenties saw a huge rise in wealth concentration, as the top 1% accumulated a significantly larger share of total wealth over the decade, rising from 35% in 1923 to almost 52% by 1928, and the top 10% wealth share peaked at 84%. But the impact of the Great Depression, and the New Deal policies implemented under Franklin Roosevelt’s Presidency, quickly saw this trend reverse.

Wealth inequality fell at a tremendous pace from 1929 until around the end of the Second World War. The loss in the value of financial assets from the collapse of the stock market and the introduction of financial regulation during the New Deal reduced the role of finance and the ability to concentrate wealth relative to the Gilded Age model of the financier-industrialist, while the development of progressive income and estate taxation made it difficult to accumulate and pass on large fortunes. Correspondingly, the share of the top 1% fell from 52% of total wealth to 29% by 1949. Their falling shares were not just accumulated by the top 10–1% either, as illustrated by Figure 4.3.1b, as the share of total wealth rose from 33% to 42%, leaving the bottom 90% with a 29% share, equal to that of the top 1%.

Following the Second World War, wealth inequality rose moderately, before falling again from the early 1960s onwards. The wealth share of the top 10% grew from around 70% to 74% in 1962, before falling in almost every year until the mid-1980s, by which point their share had dipped below 65% of total wealth. As previously described, the Reagan era of deregulation and reduced tax progressivity formed a turning point in wealth inequalities in America. The top personal income tax rate from 50% in 1986 to 28% in 1988, well below the corporate tax rate of 35%.

The rise and fall of middle-class wealth

The second key result of the analysis involves the dynamics of the wealth share of the bottom 90%. Since the bottom half of the distribution always owns close to zero net wealth, that is, when including negative wealth such as credit card and housing debt, the wealth share of the bottom 90% is therefore equal to the share of wealth owned by the middle 40% group, above the bottom 50% but below the top 10%. Within this “middle class”, the share of total wealth owned in 2012 was the same as it was 70 years earlier, despite a rise in the value of their pensions and an increase in their home ownership rates.

The share of wealth owned by the middle class began to increase from the early 1930s, and peaked in the mid-1980s. It has subsequently undergone a continuous decline, as illustrated by Figure 4.3.2. The large rise in the wealth share of the bottom 90%, from 16% in the early 1930s to 35% in the mid-1980s, was driven by the group’s accumulation of housing wealth, and to a greater extent by pensions. Pensions were almost nonexistent at the beginning of the twentieth century, but developed in the form of defined benefits plans, and then from the 1980s in the form of defined contribution plans such as Individual Retirement Accounts and the so called 401(k)s (the latter referring to a section of the United States tax code).

The declining share in the wealth share of the bottom 90% that occurred from the mid-1980s was due to a fall in two components of middle class wealth, namely the housing component (net of mortgage debt) and the fixed income component (net of non-mortgage debt). This fall was mostly the consequence of an upsurge in debt, as aggregate household debt, including mortgages, student loans, credit cards, and other debts, increased from 75% of national income in the mid-1980s to 135% in 2009. The financial crisis of 2007–2009 and the Great Recession then hit the middle class hard. The share of wealth owned by the bottom 90% collapsed between mid-2007 and mid-2008 because of the crash in housing prices, and the subsequent recovery was uneven: over 2009–2012, real wealth per family declined 0.6% per year for the bottom 90%, while it rose 7.9% per year for the top 0.1%.

Despite a reduction in debt levels in the wake of the Great Recession as the middle class sold a proportion of their assets, their debt still amounted to close to 110% of national income in 2012. This upsurge in the debt of the middle class has had a dramatic effect on middle-class wealth as approximately 90% of (non-mortgage) debt belongs to the bottom 90% of the wealth distribution, being sufficiently large to more than offset the rise in the value of their pensions. Strikingly, the average real wealth of the bottom 90% of families was no higher in 2012 than in 1986. Real average wealth of the bottom 90% rose considerably during the late 1990s tech-boom and the mid-2000s housing bubble, peaking at $143 000 in 2006, but then collapsed to about $93 800 in 2009 (at constant 2016 $), as depicted in Figure 4.3.3.

Spain has experienced an unprecedented increase in aggregate wealth over the past thirty years, predominantly due to the housing the country experienced over the last 30 years. Much has been written about this economic phenomenon, when house prices tripled between 1985 and 1991 and tripled again between 1996 and 2008,17 and the value of the stock market increased sevenfold before halving, but much less so on its distributional effects. In particular, there has been little research into which groups have benefited from this increase in wealth, how much each of these groups have benefited, how differences in wealth between groups have changed over time, whether the importance of asset categories has altered, and which factors are the source of the aforementioned changes?

Using high-quality, publicly available data, Martínez-Toledano’s recent paper18 seeks to answer these questions. The author combines tax records, national accounts and wealth surveys, as well as the capitalization method19 that is used by Saez and Zucman for the United States,20 to deliver a consistent, unified wealth distribution series for Spain between 1984 and 2013, with detailed breakdowns by age over the period 1999–2013.

The rising value of housing has fueled the growth of Spanish wealth

The Spanish personal wealth to national income ratio almost doubled between 1984 and 2014. As illustrated by Figure 4.5.1 personal wealth amounted to around 380% in the late eighties and grew to around 470% in the mid-nineties. From 1995 onwards, personal wealth started to increase more rapidly, reaching its peak at 728% of national income in 2007, before the global financial crisis. After the bubble burst in 2008, personal wealth dropped notably and continued to decrease thereafter. In 2014, the Personal Wealth to National Income ratio amounted to 646%, a level similar to the Personal Wealth to National Income ratio of years 2004 and 2005, but much higher than the ratios of the eighties and nineties, as illustrated by Figure 4.5.1.

Figure 4.5.1 also shows how the components of total net Spanish wealth have evolved over the 30-year period. The late eighties saw growth in net housing that was more than double the speed of the increase in financial assets, but this trend was reversed during the nineties as financial assets started to be accumulated at a faster pace than property, due mainly to the rise in stock prices that arose from the dot-com bubble. However, after the stock market crash of 2000, housing prices increased at a pace that surpassed even the significant growth of financial assets. The value of housing then reached its peak in 2008, after which the sizeable housing bubble that had been built up burst and the fall in housing wealth was larger than that of financial assets.

This period was also characterized by the increasing importance of net housing in the asset portfolios of households. While properties are the most important asset held by the average Spanish household between 1984 and 2014, always representing more than 40% of total household net wealth, the composition of personal wealth has not evolved homogeneously. Indeed, personal wealth has lost importance in periods when financial assets significantly increase, such as the one that preceded the dot-com bubble. The increase in the fraction of property in the total portfolio of households has also been exacerbated by the steady decrease in the fraction of unincorporated business assets, which fell from 23% in 1984 to 11% in 2014, due mainly to the relative reduction in the importance of agriculture within the Spanish economy.

The top 10% has owned more than half of Spain’s personal wealth since the mid-1980s

Table 4.5.1, displays the wealth level, threshold and shares of personal wealth for Spanish adults in 2013. On average, the net wealth per adult in Spain was approximately €144 000. However, the average wealth within the bottom 50% of the distribution was just 13% of the countrywide average, at €18 900. Cumulatively, the share of personal wealth held by the top 50% was less than 7%. Average wealth within the next 40% of the distribution was slightly over €133 000, giving the group a 37% share of personal wealth, not largely dissimilar to their population share. This left the top 10% holding over 56% of Spanish personal wealth, with an average wealth of approximately €813 000, over five-and-a-half times greater than the national average wealth and 43 times greater than the average wealth of 50% of the Spanish adult population.

The drastic differences in the shares of personal wealth reported in 2013, have remained largely unchanged throughout the preceding 29-year period. As Figure 4.5.2 shows below, the share of personal wealth held by each group has remained within a band of eight percentage points. The share of personal wealth attributable to the bottom 50% has always been very small, reaching a peak of 9% in 1992, but fell back to just over 6% in 2013, roughly equal to its level at the start of the period. The personal wealth share of the middle 40% has concentrated between 32% and 39% of total net wealth, remaining over 35% for the majority of the observed period, while the share of the top 10% has fluctuated between 53% and 61%. Notably, the top 10% wealth share dropped from the mid-eighties until the beginning of the 1990s, at the expense of the increased shares of both the middle 40% and the bottom 50% of the distribution, as house prices rose threefold across Spain. The top 10% wealth share then increased during the nineties, as the stock market grew strongly, before decreasing until the mid-2000s and increasing again until the start of the global financial crisis and burst of the housing bubble in 2008. Since then, the share of the top 10% decreased, before stabilizing at a similar level to that during the mid-nineties.

While the changes in relative assets prices have had a rather limited impact on overall wealth inequality in Spain, there are important differences in the portfolio of assets owned by different wealth groups. As shown by Figure 4.5.3, in 2013, the bottom 20% of the Spanish wealth distribution mostly owned financial assets, which largely came in the form of savings and current deposits in banks. As one move towards the center of the wealth distribution, property becomes the most dominant form of wealth (approximately 90% between the 30th and 60th percentiles). Thereafter, the dominance of financial assets within wealth portfolios grows larger as the individuals analyzed become wealthier. However, unlike the bottom 50%, bank deposits form only a minor part of financial assets for the top 10% and the top 1% of the distribution. Instead, the wealthiest Spanish adults own a combination of equities, investment funds, fixed income assets such as bonds, currency, life insurance reserves and pension funds. The same general pattern of asset composition by wealth group also applies for the period between 1984 and 2012, as can be seen in Figures 4.5.4 and 4.5.5. The only notable difference has been the falling importance of unincorporated assets over the 28-year period, which can mainly be attributed to the reduction in agricultural activity among the self-employed.

By decomposing the evolution of wealth in Spain by asset categories and by wealth group, it is possible to see how asset price movements between 1984 and 2013 affected their respective asset portfolios and shares of personal wealth. The figures within Figure 4.5.4 clearly show how the impact of the stock market boom of 2000 and the burst of the housing bubble in 2007 affected portfolios and shares of the top 1%. Reviewing the trend in the financial assets component of the wealth of the top 1%, there is an obvious spike in the value of financial assets and its dominance in their portfolio in 1999, the year preceding the dot-com crisis.

One particularity of the Spanish case relative to other rich nations is the importance of housing assets in the portfolio of households, even at the top of the distribution. This has been the case during the whole of the 29-year period analyzed, but this trend became even more striking in the years up to 2007, when the increase in the value of dwellings was largest. In Spain, the top 10% and top 1% of the wealth distribution own 26% and 8% of total net wealth in housing, respectively, whereas in France these figures are 19% and 5%, respectively.²¹

Increasingly greater sums of wealth are being passed on to the offspring of the wealthy

The detailed micro-files available in Spain from 1999 also allow Martínez-Toledano to analyze how wealth varies between different age groups, and how this has changed over time. As Figure 4.5.5 shows, average wealth has been consistently very small for those aged 20 during the 14-year period studied, at less than 10% of total wealth. Wealth exhibits a rising trend with age. At age 40, individuals own approximately 50% of average wealth whereas at age 60, they own more than 150% of average wealth. After 60, the average adult wealth declines moderately but never falling below 120%. As average wealth does not decline sharply after age 60 and remains at a level that is notably above average wealth, old-age individuals thus pass away with substantial wealth and transmit this to their offspring.

There are, however, important differences in relative wealth levels across age groups over the 1999–2013 period. Old individuals (+60) are better off and the young (20–39) worse off after the economic crisis, since the average wealth for the old relative to total average wealth is larger in 2013 than in 2001. This is consistent with the large increase in youth unemployment22 after the burst of the bubble and at the same time the stability in Social Security pension payments. When decomposing the wealth distribution series by age, it appears that wealth inequality is more pronounced for the young (20–39) than for the old (+60) and middle-old (40–59), for which wealth inequality is almost as large than for the population taken as a whole. A plausible explanation is the importance of bequests that transfer the wealth of the older generations to the younger generation. Higher transfer rates among wealthy families, combined with high youth unemployment rates and consequently a low wealth accumulation through labor income savings by the young (which would moderate wealth inequality), can explain higher inequality levels among the young than among the elderly.

The Spanish property bubble had a neutral effect on wealth inequality

The high level of disaggregation in Martínez-Toledano’s wealth distribution series also helps to explain why Spain’s housing bubble had a curiously neutral effect on the level of wealth inequality in the country. In Spain, as in many European countries, the increased ownership of property among the bottom 90%, and the significant share that housing represents in their asset portfolios, has contributed to reducing wealth inequality. Figure 4.5.6 illustrates that wealth concentration for the top 1% is approximately 10 percentage points lower between 1984 and 2013 when housing wealth is included. But moreover, the figure also shows that wealth inequality including and excluding housing followed a similar trend post 2000, confirming that the housing boom and bust had little impact on wealth inequality.

In order to understand this puzzling result, it is important to see how the composition of net housing wealth has changed over time. The fraction of total net housing owned by the top 1% increased considerably between 2005 and 2009, the years in which housing prices skyrocketed, at the expense of the proportion of homes owned by the middle 40%. This increased concentration of home ownership was principally the result of the increase in the number of secondary properties bought by the top 1%, relative to the middle 40%, and not due to relatively larger increases in the price of properties owned by the wealthiest. The ratio of the house prices of the top 10% (and top 1%) to the value of dwellings of the middle 40% remained constant between 2005 and 2009.

But if housing concentration increased at the top during the bubble and decreased thereafter, why has total wealth concentration remained virtually unchanged? One plausible explanation is that individuals within the top 1% substituted financial assets for property during the period of the housing boom, but then accumulated greater financial assets when house prices began to fall. The fraction of total financial assets held by the top 1% decreased during the boom years. This is consistent with the idea that wealthy individuals can better diversify their portfolios, and have the capabilities to invest more in risky assets, when prices are increasing—and can more easily disinvest when prices fall, to then acquire other assets.

Disparities in savings rates and returns on assets drive long-run wealth inequality

In order to understand the underlying forces driving wealth inequality dynamics in Spain, it is useful to analyse how income, savings rates and the rate of inequality have evolved between 1999 and 2012.

There are significant differences in the savings rates between wealth groups in Spain and these have changed over time, as illustrated by Figure 4.5.7a-c. These disparities reflect the high levels of wealth concentration observed in Spain, with an average savings rate of 27% of income for the top 10% over this period, compared to 10% among the middle 40% and just 1% for the bottom 50%.

Analyzing the evolution of savings rates more closely reveals one important point. The housing bubble increased the difference in saving rates between the wealthy and the less-wealthy during the boom years and reduced their stratification during the bust period. Figure 4.5.7a shows that during the years prior to the property bubble bursting, the savings rate of the top 10% remained high as they accumulated more housing, while the savings rate for the middle 40% and the bottom 50% decreased, as their accumulation of housing assets was facilitated through borrowing. After the property bubble burst, the top 10% sold some of their housing assets and started to accumulate more financial assets to compensate for the decrease in housing prices. Nonetheless, the total savings rate for the top 10% decreased during these years, likely because they needed to consume a larger fraction of their income. The middle 40% instead started to save more in order to repay their housing mortgages, and therefore the difference in saving rates across the two wealth groups was reduced. These two trends thus contributed to neutralizing wealth concentration during Spain’s tumultuous period of housing price swings.

Wealth inequality has also been amplified by the variance in the rates of return on assets owned by different wealth groups in Spain over the 1986–2012 period.²³ This finding is consistent with the large differences in the asset portfolios of Spanish wealth groups documented earlier in the chapter (Figure 4.5.1), whereby top wealth groups are more likely to own financial assets such as equity that often have higher rates of return than other assets, including deposits and housing.

Factoring in offshore wealth into the Spanish wealth distribution reveals a higher level of inequality

As is common in many other countries, official financial data in Spain fails to capture a large part of the wealth held by households abroad. Research has shown that Spanish citizens use offshore financial institutions in tax havens for their portfolios of equities, bonds, and mutual fund shares. It is estimated by Zucman²⁴ that these assets amounted to approximately €80 billion in 2012—the equivalent of 9% of households’ net financial wealth in Spain—of which three-quarters goes unrecorded. Thus, by omitting offshore wealth from the Spanish wealth distribution series, both total assets and wealth concentration are substantially underestimated.

Using data series from the Swiss National Bank, offshore wealth taxation forms and the 2012 tax amnesty, Martínez-Toledano is able to adjust her wealth distribution series for offshore assets. As illustrated by Figure 4.5.8, the value of offshore assets increased rapidly during the eighties, nineties and at the beginning of the 2000s, before stabilizing after 2007, when Spanish tax authorities became stricter with tax avoidance and evasion schemes. Unreported offshore wealth amounted to almost €150 billion in 2012, representing 8.6% of personal financial wealth. Investment funds represented 50% of total unreported offshore assets in 2012, followed by stocks, 30%, and deposits and life insurance, which made up 18% and 2%, respectively.