The Evolution of Global Inequalities: the impact on politics and the economy

Hosted by the International Inequalities Institute The Evolution of Global Inequalities: the impact on politics and the economy Professor Senior Scholar, Luxembourg Income Study Centre Visiting Presidential Professor, Graduate Centre, City University of New York Professor Mike Savage Chair, LSE Hashtag for Twitter users: #LSEBranko

Inequality in the age of globalization Spring/Summer 2017

Largely based on: 3

Brief structure of the talk Global inequality: in the past and now Technical problems of measurement How the world has changed between 1988 and 2013 [Political implications of the changes] [Kuznets waves?] Issues of justice, politics and migration

1. Global inequality: key developments

75 70 Global and US Gini over two centuries Lahoti, Jayadev, Reddy Global (LM) 65 60 Global (BM) 55 50 45 US inequality 40 35 30 1800 1850 1900 1950 2000 2050 History /the past.xls

La longue durée: From Karl Marx to Frantz Fanon and back to Marx? 80 Location Forecast 60 Gini index 40 Location Location Location 20 Class Class Class 0 History../the_past.xls 1850 2011 2050

In the long run inequality is determined by the spread of the technological revolutions: the West in the 19 th century, Asia today In the medium run global inequality is determined by: What happens to within country income distributions? Is there a catching up of poor countries? Are mean incomes of populous & large countries (China, India) growing faster or slower that the rich world?

0.75 Three concepts of inter national income inequality, 1952 2015 0.70 Global inequality 0.65 Population weighted intercountry inequality 0.60 Gini 0.55 Unweighted inter country inequality 0.50 0.45 All in 2011 PPPs 0.40 1940 1950 1960 1970 1980 1990 2000 2010 2020 Interyd\...3concepts..xls

16.0 Key developments, 1988 2011 4.5 14.0 4.0 12.0 10.0 8.0 6.0 4.0 Top 1% share (left axis) Mean to median ratio (right axis) 3.5 3.0 2.5 2.0 1.5 1.0 2.0 0.5 0.0 1988 1993 1998 2003 2008 2011 0.0

30 25 Gini and percentage of world population with income less than 1/2 global median, 1988 2011 70 69 68 Percentage of relatively poor 20 15 10 5 67 66 65 64 63 62 61 Global Gini 0 1988 1993 1998 2003 2008 2011 Axis Title % ppl under 1/2 median Gini with 2011 PPPs 60 Summary.xls

Global income distribution in 2011 with 2011 PPPs density 0.2.4.6.8 Absolute poverty Global median 10% 50% 73% Global mean 91% Median of WENAO 600 2100 5500 14600 log of annual PPP real income twoway (kdensity loginc_11_11 [w=popu] if loginc_11_11>2 & bin_year==2011, bwidth(0.2)), legend(off) title(global income distribution in 2011 with 2011 PPPs) xtitle(log of annual PPP real income) ytitle(density) xlabel(2.8"600" 3.3"2100" 3.74"5500" 4.2"14600", labsize(small) angle(90)) Using combine88_11.dta

Large gaps in mean country incomes raise two important issues Political philosophy: is the citizenship rent morally acceptable? Does global equality of opportunity matter? Global and national politics: Migration and national welfare state (will address both at the end)

Different countries and income classes in global income distribution in 2008 percentile of world income distribution 1 10 20 30 40 50 60 70 80 90 100 Russia USA China India Brazil From calcu08.dta 1 20 40 60 80 100 country percentile

Different countries and income classes in global income distribution in 2011 percentile in global income distribution 0 20 40 60 80 100 USA Russia Brazil India 1 20 50 80 100 percentile of country's income distribution India with 2011 income data Final11.dta using michele_graph.do but with india consumption replaced by india income

Why international aid is unlikely to involve regressive transfers? percentile of world income distribution 1 10 20 30 40 50 60 70 80 90 100 Netherlands Mali Tanzania Guinea Madagascar 1 20 40 60 80 100 percentile of country's income distribution

2. Technical issues in the measurement of global inequality

Three important technical issues in the measurement of global inequality The ever changing PPPs in particular for populous countries like China and India The increasing discrepancy between GDP per capita and HS means, or more importantly consumption per capita and HS means Inadequate coverage of top 1% (related also to the previous point)

The issue of PPPs

The effect of the new PPPs on countries GDP per capita gain compared to 2005 ipc--normalized by the us level -50 0 50 100 150 SAU ZMB SDN JOR GHA IDN MNG SUR OMN KWT PAK EGY NPLBGD FJI AZEKAZ QAT YEM CIV LAO CPV DZA THA MDG LKA MAC VNMPHLGTM NER BRN MLI MAR HTITCD COG VENRUS GNQ MYS ARE TGO KEN MRT IND MDV LSO BDI SLE UGA KGZ NGAMDA AGONAM BRA GIN CMR SWZ CHN KHM BTN UKR ETH BLR TJK NIC BOL TUNMKD GNB RWA BFA BEN SEN GEO PRY MNE ARM BIH BGR TUR LVA CAF MWI HND SLV BLZECU DOM PER HUN SGP COL MEX URY CHL TTO SRB ZAF LTU EST MUS HRV POL TZA JAM CRIGAB SVK MLT ITA DNKCHE NOR LUX NZL PAN PRT GRCESP FRA FIN TWN BEL DEU SWE IRLUSA MOZ DJI ALB CZE SVN ISR ISL AUT AUS CAN NLD JPN HKG LBR KOR GBR GMB BWA CYP COM gdppc in 2011ppp BHS 50000 100000 150000 C:\Branko\worldyd\ppp\2011_icp\define

Country The effect of new PPPs GDP per capita increase (in %) GDP per capita increase populationweighted (in %) Indonesia 90 Pakistan 66 Russia 35 India 26 China 17 Africa 23 32 Asia 48 33 Latin America 13 17 Eastern Europe 16 24 WENAO 3 2

Use of 2011 PPPs reduces global inequality by about 3 Gini points but leaves the trends the same 74.0 72.0 70.0 68.0 66.0 64.0 62.0 60.0 58.0 Using summary_data.xls 1988 1993 1998 2003 2008 2011 Gini with 2011 PPPs Gini with 2005 PPPs

The gap between national accounts and household surveys

Global Gini with different definitions of income 74 72 70 68 Step 2 Step 1 HH survey NA consumption 66 64 GDP per capita 62 60 Summary_data.xls 1988 1993 1998 2003 2008

Step 1 driven by low consumption shares in China and India (although on an unweighted base C/GDP decreases with GDP) C/GDP from national accounts in year 2008 share of consumption in GDP.2.4.6.8 1 1.2 India China USA 1000 10000 50000 GDP per capita in ppp twoway scatter cons_gdp gdpppp if group==1 & cons_gdp<1.4 [w=totpop], xscale(log) xtitle(gdp per capita in ppp) xlabel(1000 10000 50000) ytitle(share of consumption in GDP) title(c/gdp from national accounts in year 2008) using final08,dta

Step 2. No clear (weighted) relationship between survey capture and NA consumption survey mean over NA consumption.2.4.6.8 1 1.2 survey mean/consumption from national account in year 2008 India China USA 1000 10000 50000 GDP per capita in ppp twoway scatter scale2 gdpppp if group==1 & scale2<1.5 [w=totpop], xscale(log) xtitle(gdp per capita in ppp) xlabel(1000 10000 50000) ytitle(survey mean over NA consumption) title(survey mean/consumption from national account in year 2008)

The issue of top underestimation

Rising NAC/HS gap and top underestimation If these two problems are really just one & the same problem. Assign the entire positive (NA consumption HS mean) gap to national top deciles Use Pareto interpolation to elongate the distribution No a priori guarantee that global Gini will increase

20 18 Top 1% share in US: Comparison between WTID fiscal data and factor income from LIS (both run across households/fiscal units; K gains excluded) WTID data 16 14 12 10 LIS CPS data 8 6 4 2 0 1975 1980 1985 1990 1995 2000 2005 2010 2015 usa07_13.xls

But the rising gap between fiscal and HS income is not universal 14.0 Top 1% share Norway: Comparison between WTID fiscal data and factor income from LIS (both run across households/fiscal units; K gains excluded) 12.0 WTID data 10.0 8.0 6.0 LIS data 4.0 2.0 0.0 1975 1980 1985 1990 1995 2000 2005 2010 2015

With full adjustment (allocation to the top 10% + Pareto) Gini decline almost vanishes 80 78 76 74 72 Top heavy allocation of the gap + Pareto adjustment Survey data only 70 68 66 64 1988 1993 1998 2003 2008 Summary_data.xls

3. How has the world changed between the fall of the Berlin Wall and the Great Recession [based on joint work with Christoph Lakner]

Number of surveys 1988 1993 1998 2002 2005 2008 2011 Africa 14 30 24 29 32 23 30 Asia 19 26 28 26 23 27 22 E.Europe 27 22 27 25 27 27 24 LAC 19 20 22 21 18 18 18 WENAO 23 23 21 21 22 23 21 World 102 121 122 122 122 118 115

Population coverage 1988 1993 1998 2002 2005 2008 2011 Africa 48 76 67 77 78 78 70 Asia 93 95 94 96 94 98 96 E.Europe 99 95 100 97 93 92 87 LAC 87 92 93 96 96 97 97 WENAO 92 95 97 99 99 97 96 World 87 92 92 94 93 94 92 Non triviality of the omitted countries

GDI (US dollar) coverage 1988 1993 1998 2002 2005 2008 2011 Africa 49 85 71 71 70 71 63 Asia 94 93 96 95 90 93 83 E. Europe 99 96 100 99 99 98 94 LAC 90 93 95 95 98 98 94 WENAO 99 96 96 100 100 97 95 World 96 95 96 98 97 95 90

Real PPP income change (in percent) Real income growth at various percentiles of global income distribution, 1988 2008 (in 2005 PPPs) 80 70 60 50 40 30 20 10 $PPP2 X China s middle class $PPP4.5 $PPP12 $PPP 180 X US lower middle class 0 0 20 40 60 80 100 Percentile of global income distribution From twenty_years\final\summary_data Estimated at mean over mean

Parts of the distribution that gained the most are dominantly from Asia, parts that stagnated are mostly from mature economies Cumulative growth rate (%) 90 80 70 60 50 40 30 20 Quasi-non-anonymous growth incidence curve (1988-2008) population-weighted, including population distribution in base-year 0.2.4.6.8 1 Normalised rank in the 1988 global income distribution Asia Mature economies GIC 100 90 80 70 60 50 40 30 20 10 0 Share of region in ventile population (%) Solid line shows predicted value from kernel-weighted local polynomial regression (bw=0.05, epanechnikov, cube polynomial). Only countries observed in 1988 & 2008 (N=63) included. From analysis horizontal quasinonanon gic pop 2do

Quasi non anonymous growth between 1988 and 2008: real absolute per capita gains at different fractiles of the 1988 distribution Absolute per capita real income gain between 1988 and 2008 25000 20000 15000 10000 5000 0 22891 7190 140 217 236 443 466 799 1246 1163 2218 3168 D1 D2 D3 D4 D5 D6 D7 D8 D9 V19 P95 99 P100 Decile/fractile of 1988 global income distribution

140 Real income growth over 1988 2008 and 1988 2011 (based on 2011 PPPs) Cumulative real per capita growth in % between 1988 and 2008 120 100 80 60 40 20 1988 2011 1988 2008 0 0 10 20 30 40 50 60 70 80 90 100 Percentile of global income distribution

Global income distributions in 1988 and 2011 Figure 3. Global income dstribution in 1988 and 2011 density 0.2.4.6.8 1988 2011 Emerging global middle class between $3 and $16 300 1000 3000 10000 log of annual PPP real income 50000 twoway (kdensity loginc_11_11 [w=popu] if loginc_11_11>2 & bin_year==1988, bwidth(0.14) title("figure 3. Global income distribution in 1988 and 2011")) (kdensity loginc_11_11 [w=popu] if loginc_11_11>2 & bin_year==2011, bwidth(0.2)), legend(off) xtitle(log of annual PPP real income) ytitle(density) text(0.78 2.5 "1988") text(0.65 3.5 "2011") xlabel(2.477"300" 3"1000" 3.477"3000" 4"10000" Branko 4.699"50000", Milanovic labsize(small) angle(90)) Using Branko\Income_inequality\final11\combine88_08_11_new.dta

From Christoph Lakner

4. Political implications

The contradiction of inequality changes during Globalization II Most countries displayed an upward sloping GIC (US, China, India urban, Indonesia ) Perception that the rich are doing better than anybody else (true) But growth rates of countries are uneven; those that grew the fastest were in the lower middle of global income distribution, and they were also most populous This led to the elephant shaped global GIC and decreasing global inequality

The issues Are growth (1) along the entire Chinese income distribution and (2) stagnation around the median in the rich world as well as stagnation across most of income distribution in E. Europe and LAC, related? In other words, is the hump in middle related to the dip around the 70 80 th percentile? Marching of China and India through the ranks reduces global inequality and the importance of the betweencountry component in global inequality But it might cause increases in within national inequalities (thus offsetting global inequality decline) Can democracy survive if rich countries middle classes are hollowed out?

Back to Mandeville Can something that is bad nationally (increased inequality) be good globally (decreased inequality)? Can national vices produce global virtue?

Political implications Possible crowding out of national middle classes, and the creation of a global one But the middle class is presumably a force for stability when there is a political community. There is no political community at the global level. What does global middle class mean? Would global middle class create a global polity? Or, global plutocracy: in the longer term, reversal to the pre World War I situation

Are we at the end of capitalism s long el periodo especial or going upward the second modern era Kuznets curve? Three challengers to global capitalism were beaten off in the 20 th century: depression (by reinventing gov t), war (by marshalling resources), Communism (through Welfare State) Neither of these threats is any longer present; so is this the reason capitalism is becoming more unequal? Or is the period after 1980, the second modern era Kuznets curve driven by the technological revolution and globalization?

Focus on point B of the elephant graph (income stagnation and erosion of the middle class in advanced economies)

Income share of the middle four deciles 1980-2013 in percent 28 30 32 34 USA 28 30 32 34 UK 1980 1990 2000 2010 2020 year 1980 1990 2000 2010 2020 year 28 30 32 34 Germany 28 30 32 34 Canada 1980 1990 2000 2010 2020 year 1980 1990 2000 2010 2020 year c:\branko\voter\dofils\define_variables using data_voter_checked.dta

Percentage of population considered middle class in early 1980s and 2013 Finland 43 50 Netherlands 42 45 UK 33 40 Germany 37 40 Canada * Spain 30 34 36 35 USA 27 32 0 10 20 30 40 50 60 around 2013 The middle class defined as population with income between +/ 25% of national median income (all in per capita basis; disposable income; LIS data)

20,000 US real median after tax household per capita income 1979 2013 18,000 16,000 14,000 12,000 10,000 8,000 6,000 4,000 Average annual growth rate over the entire period: 0.5% Since 2000, zero. 2,000 0 1979 1986 1991 1994 1997 2000 2004 2007 2010 2013

5. How to think of within national inequalities: Introducing the Kuznets waves

Kuznets waves defined Kuznets saw just one curve. We now know there may be many more. Distinguish the waves in pre industrial and modern societies (those with sustained increase in mean income) Kuznets waves in pre industrial societies are visible when plotted against time only (because mean income is stagnant) Kuznets cycles in industrial societies are visible when plotted against income per capita=> proxy for structural changes Inequality waves are too complex for formal modelling => need to use inductive reasoning and analytic narrative The waves in modern era reflect economic forces of technological innovation and structural transformation. But also wars and policy changes. 53

Malign and benign forces reducing inequality (downward portion of the Kuznets wave) Societies with stagnant mean income Societies with a rising mean income Malign Idiosyncratic events: wars (though destruction), epidemics, civil conflict Wars (through destruction and higher taxation: War and Welfare), civil conflict Benign Cultural and ideological (e.g. Christianity?) Widespread education (reflecting changing returns) Social pressure through politics (socialism, trade unions) Aging (demand for social protection) Low skill biased TC Cultural and ideological (pay norms?) 54

140 120 100 80 60 40 20 0 Cyclical nature of the Kuznets curve: Land rental/wage ratio over the long-term in Spain, 1282-1842 Napoleonic wars Wool and wine production, rising demand for land, commercial society Wars, decline of wool exports Plague 1282 1294 1306 1318 1330 1342 1354 1366 1378 1390 1402 1414 1426 1438 1450 1462 1474 1486 1498 1510 1522 1534 1546 1558 1570 1582 1594 1606 1618 1630 1642 1654 1666 1678 1690 1702 1714 1726 1738 1750 1762 1774 1786 1798 1810 1822 1834 1846 From Prados de la Escosura & Alvarez-Nogal, The rise and fall of Spain 800-1850 Land rental/wage ratio

Kuznets curve here? No. GDP per capita and rent wage ratio: Spain 1325 1840 140 120 Lad rent/wage ratio 100 80 60 40 Land/wage 20 0 0 10 20 30 40 50 60 70 80 90 100 GDP per capita (1859 100) From Prados de la Escosura & Alvarez Nogal, The rise and fall of Spain 800 1850 56

Kuznets relationship for the UK, 1688 2014 60 1867 50 1913 Gini of disposable per capita income 40 30 20 1688 1962 1978 1993 2013 10 0 0 5000 10000 15000 20000 25000 30000 GDP per capita (in 1990 international dollars; Maddison)

Kuznets relationship for the United States, 1774 2013 60 50 1860 1933 1774 1929 2013 Gini of disposable per capita income 40 30 20 1947 1979 10 0 0 5000 10000 15000 20000 25000 30000 35000 GDP per capita (in 1990 international dollars; Maddison)

The Kuznets relationship for the Netherlands, 1561 2010 70 1732 60 1808 50 1561 1914 1962 40 Gini 30 1982 2010 20 10 0 0 5000 10000 15000 20000 25000 30000 GDP per capita (in 1990 international dollars)

Downswing of Kuznets first wave and upswing of the second Kuznets wave in advanced economies Level of maximum inequality (peak of Wave 1) Gini points (year) Level of minimum inequality (trough of Wave 1) (year) Approximate number of years of downswing of the Kuznets wave Reduction in inequality (Gini points) GDP increased (how many times) during the downswing The second Kuznets wave (increase in Gini points) United States 51 (1933) 35 (1979) 50 16 4 Strong (+8) UK 57 (1867) 27 (1978) 110 30 >4 Strong (+11) Spain 53 (1918) 31 (1985) 70 22 <5 Modest (+3) Italy 51 (1851) 30 (1983) 120 21 <9 Strong (+5) Japan 55 (1937) 31 (1981) 45 24 6 Modest (+1) Netherlands 61 (1732) 28 (1982) 250 33 7 Modest(+2) Table2_data.xls 60

0.35 Urban Gini in China: 1981 2014 (based on official household surveys) 0.3 0.25 Urban Gini 0.2 0.15 0.1 0.05 0 Year

Where are now China and the US? Gini First Kuznets wave Second Kuznets wave China 2013 United States 2013 GDP per capita

What might drive the 2 nd Kuznets cycle down? Progressive political change (endogenous: political demand) Dissipation of innovation rents Low skilled biased technological progress (endogenous) Reduced gap in education (but it is not a silver bullet) Global income convergence: Chinese wages catch up with American wages: the hollowing out process stops Note that all are all endogenous 63

6. Issues of justice and politics 1. Citizenship rent 2. Migration and national welfare state 3. Hollowing out of the rich countries middle classes

Global inequality of opportunity Regressing (log) average incomes of 118 countries percentiles (11,800 data points) against country dummies explains 77% of variability of income percentiles Where you live is the most important determinant of your income; for 97% of people in the world: birth=citizenship. Citizenship rent.

Is citizenship a rent? If most of our income is determined by citizenship, then there is little equality of opportunity globally and citizenship is a rent (unrelated to individual desert, effort) Key issue: Is global equality of opportunity something that we ought to be concerned or not? Does national self determination dispenses with the need to worry about GEO?

The logic of the argument Citizenship is a morally arbitrary circumstance, independent of individual effort It can be regarded as a rent (shared by all members of a community) Are citizenship rents globally acceptable or not? Political philosophy arguments pro (social contract; statist theory; self determination) and contra (cosmopolitan approach)

Rawls views on inter generational transmission of wealth Group Intergenerational transmission of collectively acquired wealth Argument Policy Family Not acceptable Or at least to be limited Threatens equality of citizens Moderate to very high inheritance tax Nation Acceptable Affirms national selfdetermination (moral hazard) International aid

The Rawlsian world For Rawls, global optimum distribution of income is simply a sum of national optimal income distributions Why Rawlsian world will remain unequal?

Global inequality in Real World, Rawlsian World, Convergence World and Shangri La World (Theil 0; year 2011) Individual incomes within country All equal Mean country incomes Different (as now) All equal 0 23 (all mean incomes equalized; all country Ginis as now) Different (as now) 54 (all country Theils=0; all mean incomes as now) 77

Conclusion Working on equalization of within national inequalities will not be sufficient to significantly reduce global inequality Faster growth of poorer countries is key and also

Migration.

Migration: a different way to reduce global inequality and citizenship rent How to view development: Development is increased income for poor people regardless of where they live, in their countries of birth or elsewhere Migration and LDC growth thus become two equivalent instruments for development

Growing inter country income differences and migration: Key seven borders today

The logic of the migration argument Population in rich countries enjoys the citizenship premium They are unwilling to share, and thus possibly reduce (at least locally ) this premium with migrants Currently, the premium is full or 0 because citizenship is (in terms of rights as well as financially) a binary variable Introduce various levels of citizenship (tax discrimination of migrants; obligation to return; no family etc.) to reduce the premium Temporary work Doing this should make native population more acceptant of migrants

Trade off between citizenship rights and extent of migration Full citizen rights Seasonal workers (almost 0 rights) 0 13% of world population* Migration flow * People who would like to migrate according to a world wide Gallup poll

Political issue: Global vs. national level Our income and employment is increasingly determined by global forces But political decision making still takes place at the level of the nation state If stagnation of income of rich countries middle classes continues, will they continue to support globalization? Two dangers: populism and plutocracy To avert both, need for within national redistributions: those who lose have to be helped

Final conclusion To reduce global inequality: fast growth of poor countries + migration To have migration, discriminate the migrants To preserve good aspects of globalization: reduced inequality within rich countries via equalization of human and financial assets (i.e. focus on pre redistribution)