This article provides detailed insights into the distributional effects of the 2007-2008 crisis and aftermath government policies, in Portugal. It sheds new light on the distributional consequences of aggregate crises and crisis-coping policies, providing useful information for improving their design.
- The 2007-2008 crisis and aftermath government policies had significant impacts on the Portuguese income distribution, reflecting markedly different developments over two periods;
- Between 2007 and 2009, stimulus packages determined important income gains for the bottom of the distribution and a decrease in income inequality;
- Between 2010 and 2013, the sovereign debt crisis and austerity measures took a toll on the incomes of Portuguese households, particularly those at the bottom and top of the distribution, leading to an increase in income inequality;
- The Portuguese experience has important policy implications. It shows that government policies can crucially influence the heterogeneity of the effects of a crisis on households' incomes and highlights the need to strike a balance between stabilising aggregate outcomes and shielding households from extreme and unequal income changes in a crisis aftermath.
The 2007-2008 crisis brought about significant changes in the income distribution of several European countries. Beyond the direct effects of the crisis, many countries were impacted by fiscal stimulus packages implemented in the early stages of the crisis and by fiscal consolidation measures implemented later on, following the start of the sovereign debt crisis. Understanding the contribution of these different aspects to the overall changes in the income distribution is key for policy makers, as it gives crucial information for the optimal design and implementation of crisis-coping policies.
In a recent paper, Sologon, Almeida and Van Kerm (2019) propose a new method that enables a thorough assessment of these issues. They apply it to the study of one particularly interesting case study: Portugal between 2007 and 2013, including the effects of the Economic Adjustment Program (EAP), signed between the Portuguese government and the Troika (composed by the International Monetary Fund, the European Central Bank and the European Commission (EC)). This article presents the main results and policy implications of this analysis, which gives a net contribution to the debate on the effects of implementing stimulus and consolidation programs in times of crisis.
The 2007-2008 crisis and aftermath policies in Portugal
The financial and economic turmoil that emerged in the global economy following the outbreak of the 2007-2008 crisis in the US first hit Portugal through the banking sector, uncovering the fragilities of two private banks, BPN and BPP, caused by years of bad management. On the grounds of avoiding a financial crisis, the Portuguese government, led by José Socrates, decided to give the two banks a bailout. This put a significant strain on the country's public finances, in a context where the deficit and debt to GDP ratio were already high (the deficit was 3% of GDP and the debt to GDP ratio was close to 70% in 2007).
At the same time, the government incurred in fiscal stimulus measures to cushion the effect of the crisis, which further contributed to a deterioration of the country's fiscal stance. These stimulus measures adopted in the early stages of the crisis were part of a European wide recovery strategy, following the adoption of the European Economic Recovery Plan (EERP) by the EC at the end of 2008. The Portuguese government took significant actions to respond to the requests included in the EERP, including an increase in government guarantees on bank deposits, the nationalization of BPN and the Investment and Employment Initiative (IEI). Beyond the efforts required by the EERP, the government took additional measures, in an anticipation of elections by the end of 2009. Among these, there were a raise for public employees, a decrease in the standard VAT rate, and measures to strengthen unemployment insurance and social assistance schemes, including an increase in the duration and amount of unemployment benefits and an easing of the qualifying conditions for social assistance recipiency.
Financial markets became anxious about the health of the Portuguese economy and public finances and the Republic's risk premium on government bonds started to increase. This gave rise to a wave of downgrades of the country's sovereign bond rating, starting with the cut by Moody's from Aa2 to A1, in July 2010. The concerns about the (un)sustainability of public finances and the pressure from financial markets led the government to abandon the fiscal stimulus strategy it was following and adopt several fiscal consolidation measures instead, starting in September 2010. These measures were not enough to ease the tensions, which culminated in a political crisis that led to José Socrates's resignation in March 2011 and led to a request in April 2011 for financial assistance from international institutions. In May 2011, a Memorandum of Understanding was signed by the resigning prime-minister and the Troika, defining a set of austerity measures to be taken, in the context of an EAP aimed at promoting the return of the economy to a balanced economic and fiscal situation. A new government headed by Pedro Passos Coelho came into functions in June 2011. The austerity policy lasted until the end of Mr. Coelho's government in November 2015, with most measures being taken between 2012 and 2013.
The consolidation measures taken from 2010 onwards were sizeable and across-the-board. On the expenditure side, reductions and restrictions on family benefits and social assistance were introduced, including child benefits and minimum income. Unemployment insurance and assistance was also weakened, through a reduction in the duration, hardening of eligibility conditions, and cut in the rates. Furthermore, there were massive cuts in public sector pay, a freezing of promotions in public careers, cuts and freezes in pensions and changes in pension indexation rules determining a lower updating. On the revenue side, there were increases in the rates of both direct and indirect taxes and the introduction of an additional income tax rate for top earners. Furthermore, there was a decrease in the amount of tax benefits allowed to be deducted from tax liabilities. Besides measures directly aimed at reducing the public deficit, there were measures targeting an increase in labour market flexibility, including a weakening of collective bargaining agreements, cuts in severance payments, and a simplification of layoff procedures. Furthermore, there was an induced retirement of civil servants, which was reinforced by a rule of only one admission for every two exits in the public sector.
The joint effects of the crisis and austerity measures taken from 2010 onwards led to a substantial worsening of the Portuguese macroeconomic and labour market context, including a severe drop in economic activity and soaring unemployment. There was a recession in four years, with real GDP growth being -3% in 2009, -1.8% in 2011, -4% in 2012 and -1.1% in 2013. The unemployment rate rose steadily between 2008 and 2013, from 8% to 16%. Furthermore, the fiscal position deteriorated considerably, with the public deficit reaching -10% of GDP in 2009, -11% of GDP in 2010 ,and exhibiting high values until 2014 and the debt to GDP ratio increasing from around 70% in 2007 to 130% in 2014.
The evolution of the Portuguese income distribution following the crisis
To assess the changes in the Portuguese disposable income distribution following the crisis, we compute the differences between the 2007, 2009 and 2013 full distributions, given by Pen's parades. Figure 1 shows the pairwise differences between these distributions, as a percentage of the 2007 distribution . It is clear that there were two markedly different sub-periods. The years between 2007 and 2009 brought income gains for almost all percentiles, particularly those at the bottom of the distribution (the poorest 30% experienced increases between 10% and 25%). This is in sharp contrast with the years between 2010 and 2013, where all percentiles experienced income losses, particularly poorer ones (those below the 30th percentile suffered losses between 10% and 35%). Looking at the 2007-2013 period as a whole, we see that for lower and middle-upper percentiles the gains observed between 2007 and 2009 were more than offset by the losses observed between 2010 and 2013, while for the middle of the distribution (percentiles 20 to 60) disposable income was left relatively unchanged.
Figure 1: Changes in the distribution of disposable income
Sources: Sologon, Almeida, Van Kerm (2019), BSI Economics
Besides looking at changes in percentiles of the income distribution, it is also informative to assess changes in income inequality and redistribution. Table 1 presents some indicators, based on the Gini index . It shows that, as for the income distribution, the evolution of income inequality encompassed two very distinct periods. While the period between 2007 and 2009 was one of decreasing inequality, particularly for disposable income, the period between 2010 and 2013 was one of increasing inequality, particularly for market income. Furthermore, the tax and transfer system was a crucial determinant of the level of disposable income inequality. Indeed, in each of the years considered, absolute redistribution was substantial, equal to more than one third of market income inequality, on average. Finally, one can see that the system was also crucial at taming the rise in market income inequality that occurred between 2007 and 2013, with redistribution increasing in both the 2007-2009 and the 2010-2013 sub-periods, by approximately 3 Gini points in each. Taking the 2007-2013 period as a whole, we see that market income inequality was essentially driven by the increase that occurred in the second sub-period, with the Gini increasing by 3.6 points, whereas disposable income inequality mostly reflected the decrease that occurred in the first sub-period, with the Gini decreasing by 2.1 points.
Table 1: Changes in income inequality and redistribution
Sources: Sologon, Almeida, Van Kerm (2019), BSI Economics
The drivers of post-crisis changes in the Portuguese income distribution
What were the drivers of the distributional changes identified in the previous section? In their paper, Sologon, Almeida and Van Kerm (2019) isolate the contributions of four main factors: (i) labour market structure; (ii) returns; (iii) demographic composition; and (iv) tax-benefit system . As previously discussed, it is crucial to distinguish between two sub-periods: 2007 to 2009, when the crisis had not yet hit the country in full force and stimulus measures were adopted; 2010 to 2013, when the crisis had its most profound recessive effects and the country was subject to a fiscal consolidation program.
Starting with the 2007-2009 period, three main conclusions can be drawn. First, changes in the labour market structure did not significantly impact households' incomes, except for poorer households, for whom it determined minor income losses. This result is likely to reflect three main aspects: the fact that the crisis hit the economy the hardest only from mid-2009 onwards; the effectiveness of stimulus measures in preventing large employment losses in the immediate aftermath of the crisis; the role of automatic stabilisers in cushioning income losses for those who did lose their job.
Second, changes in returns determined significant income gains across the whole income distribution, but particularly for households at the middle and at the top, having a disequalising effect. This can be fully linked to changes in labour incomes, as the increases in wages and salaries promoted by fiscal stimulus measures benefited relatively more households at the upper half of the income distribution.
Third, changes in the tax-benefit system determined important income gains for households at the bottom half of the distribution, having an equalising effect. This is in line with the reinforcement of the social protection system in the context of the adopted fiscal stimulus package. Although discretionary changes in the tax-benefit system were the biggest contributor for redistribution, automatic stabilisers also had a meaningful role, counteracting some of the market led increase in inequality.
Moving to the 2010-2013 period, results are markedly different. First, changes in the labour market determined income losses for bottom and middle income households, having a disequalising effect. This is likely to reflect two main aspects: the direct recessive impacts of the crisis, which started to be felt in 2009 and embodied more pronounced employment losses for lower skilled workers; the contractionary effects of austerity measures taken from 2010 onwards. However, the losses were not particularly marked, which points to a strong role of automatic stabilisers in cushioning disposable income from the effects of employment losses, particularly through unemployment benefits.
Second, changes in returns were again a crucial determinant, but this time implied significant losses across the whole income distribution, particularly for middle and top income households, having a disequalising effect. This is in line with the strong decrease in wages and salaries observed from 2010 onwards, particularly in the public sector, following the cuts in civil servants' pay introduced by the EAP.
Third, changes in the tax-benefit system were also crucial, implying a reduction in the incomes of all quantiles, particularly for lower ones and to a smaller extent for upper ones. Again, this is consistent with the developments that occurred in the context of the austerity measures taken from 2010 onwards. The substantial weakening of benefits led to important losses for the bottom of the income distribution, while the increases in taxes and cuts in pensions affected relatively more middle and higher income households, which is likely to explain the observed income losses for this group. Taken as a whole, discretionary changes in the tax and transfer system were equalising, as the losses for middle and top income households outweighed the losses for bottom income households. Automatic stabilisers once again played an important role, mitigating the disequalising effect of changes in the labour market structure and reinforcing the equalising effect of changes in returns.
Fourth, changes in the demographic composition also had a non-negligible effect, determining income gains for households at the middle and top of the distribution and an increase in inequality. This is likely to result from a composition effect, reflecting an increase in the share of highly educated and older workers, due to the exit from the labour market of lower skilled and younger workers (as lower skilled workers were strongly affected by employment losses, and a fraction of younger workers emigrated searching for employment opportunities).
Considering the 2007-2013 period as a whole, it is clearthat the observed changes are a mix of the contrasting developments in each of the two sub-periods. First, changes in the labour market structure contributed to income losses for households at the 40 lowest percentiles and an increase in income inequality. This reflected the increase in unemployment that occurred over the period but particularly between 2009 and 2013, which penalised more heavily households at the bottom of the distribution.
Second, changes in returns led to income losses for all households particularly for middle-upper income ones. The substantial cuts in wages and salaries implemented from 2010 onwards as part of the austerity packages more than compensated the gains induced by the stimulus measures in the immediate aftermath of the crisis.
Third, changes in the tax-benefit system implied income gains for households between the 10th and 30th percentile, but losses for all other percentiles, particularly those at the very bottom and top of the distribution. For these households, the losses suffered in the second sub-period following the important cuts in benefits and pensions and tax increases more than compensated the gains obtained in the first sub-period from the increase in benefits. Taken together, these developments contributed to an increase in redistribution and a decrease in inequality. Automatic stabilisers also played an important role, absorbing part of the losses in market income and rise in market income inequality.
Finally, changes in the demographic composition led to income gains for higher income households and a rise in inequality, reflecting an increase in the share of highly educated and older workers between 2009 and 2013, in a context where lower-skilled and younger workers may have exited the labour market due to important employment losses and emigration.
The evolution of the Portuguese income distribution following the 2007-2008 crisis was marked by two very distinct periods. Between 2007 and 2009, the adoption of fiscal stimulus measures determined income gains for households across the whole distribution, particularly those at the bottom half, contributing to an increase in redistribution and a reduction in disposable income inequality. Between 2010 and 2013, the combined effects of the sovereign debt crisis and austerity measures determined income losses for households in all percentiles, especially for those at the bottom and at the very top, implying a rise in disposable income inequality, despite an increase in redistribution.
Several lessons can be drawn from the post 2007-2008 crisis Portuguese experience. First, aggregate crises are likely to have not only significant aggregate impacts but also important distributional consequences, hurting relatively more low income households. Second, implementing fiscal stimulus packages following a crisis can be effective not only at stabilising aggregate outcomes but also at rendering the income distribution more equal. Conversely, the implementation of austerity measures may reinforce income losses induced by the contractionary effects of the crisis and reduce protection of the poorest. Third, beyond discretionary changes in tax-benefit rules, automatic stabilisers may be crucial at minimising income losses and preventing a rise in income inequality following a crisis.
The bottom line is that government policies can crucially influence the heterogeneity of the effects of a crisis on households' incomes, determining important income gains or losses for different income groups. When facing rising unemployment, decreasing aggregate activity and growing budget deficits, following a crisis, governments need to carefully consider the distributional impacts of their policy choices, and search for a balance between stabilising aggregate outcomes and shielding households from extreme and unequal income changes.
Bourgignon, François, and Ferreira, Francisco, and Leite, Philippe (2008), Beyond Oaxaca-Blinder: accounting for differences in household income distributions, Journal of Economic Inequality, 6 (2).
Sologon, Denisa, and Almeida, Vanda, and Van Kerm, Phillippe (2019), Accounting for the distributional effects of the 2007-2008 crisis and the Economic Adjustment Program in Portugal, LISER Working Papers, No. 2019-05.
Sologon, Denisa, and Van Kerm, Philippe, Li, Jinjing, and O'Donoghue, Cathal (2018), Accounting for differences in income inequality across countries: Ireland and the United Kingdom, LISER Working Papers, No. 2018-01.
 For each percentile, the change over the whole 2007-2013 period adds up to the sum of the changes in each of the two sub-periods, 2007-2009 and 2009-2013.
 The Gini of market (disposable) income measures inequality before (after) government transfers are added and taxes and social insurance contributions are deducted. Absolute redistribution is the difference between the two Ginis, measuring the redistributive power of the tax and transfer system. Relative redistribution is the ratio of absolute redistribution and the Gini of market income, giving the fraction of market income inequality that is reduced due to the tax and transfer system.
 Here we focus on providing only a summary of the main results, but all the figures and tables, together with a more detailed discussion of the results, can be found in the paper, for the interested reader.
 Automatic stabilisers are elements of the tax and transfer systems that temper the economy when it overheats and stimulate the economy when it slumps, without direct intervention by policymakers.