U.S. Uncertainty and Income Inequality in Gulf Cooperation Council Countries

Abstract

Economic and institutional determinants of income inequality had been robustly examined. However, the impact of uncertainty, which is more volatile than other economic indicators, on income distribution, has been largely understudied within the literature. Using new data on economic policy (EPU) uncertainty constructed by Baker et al. (2013), and the World Inequality Database (WID), this article examines the impact of U.S. EPU on income inequality in Gulf Cooperation Council (GCC) countries since 1980, using an error correction model (ECM). It is found that U.S. EPU dynamics can widen the income inequality gap in GCC.

JEL Codes

D31, D63, D80

Keywords

income inequality economic policy uncertainty Gulf Cooperation Council

Introduction

Income is becoming highly concentrated. For example, the top 10% of earners in the U.S. seized 45.5% of the total income (WID, 2019). Although not as in the US, in Europe, income highly concentrated as well. In 2020, the top 10% in Germany and U.K captured more than a third of the income share (WID, 2019). The problem of inequality is exacerbated in Gulf Cooperation Council Countries (GCC).¹ The top decile in the United Arab Emirates (UAE) and Kingdom of Saudi Arabia (KSA) controlled 56.7% and 54% of the income share, respectively (WID, 2019). One of the obstacles that explain why economists have not investigated the determinants of income distribution in GCC countries is the lack of data. However, the World Inequality Database (WID) that has been constructed recently provides rich comparable time-series data for income distribution in several developed and developing economies.

The topic of income inequality spillovers has long been preoccupied with investigating its effects on macroeconomic variables or vice versa. Robust mixed debates have shown how income distribution and economic growth are related (Alesina & Rodrick, 1994; Banerjee and Duflo, 2003; Fawaz et al., 2014). Moreover, the role that developments in the financial sector can play in shaping income inequality has also been scrutinized (Beck, Demirgüç-Kunt, & Levine, 2007; De Haan & Sturm, 2017; Demirgüç-Kunt & Levine, 2009). The quest got extended to include investigating the role that central banks play in shaping the dynamics of inequality (Colciago, Samarina, & de Haan, 2019; Samarina and Nguyen, 2019). In contrast, the literature has spent less time-if any-on examining the nexus between economic policy uncertainty (EPU) and income inequality.

This is puzzling given the fact that EPU is a major trigger of the fluctuations in the business cycle, and volatility in output can widen income inequality across countries (Fawaz, Rahnamamoghadam, & Valcarcel, 2012). Spikes in uncertainty drive-up levels of volatility in investment and employment in the sectors that are more prone to shocks in uncertainty, including infrastructure, health care, and defense (Baker, Bloom, & Davis, 2016).

Interests in studying uncertainty and its impact on the economy have intensified since the great recession. The Federal Open Market Committee (FOMC, 2009) and the International Monetary Fund (IMF, 2013) emphasized that ambiguous fiscal policies in U.S. and Europe have contributed heavily to economic imbalances in 2008–09.

Since the U.S. is a large open economy, it could directly impact its trade partners. (Colombo, 2013) argued that U.S. EPU has far-reaching repercussions on European industrial production and prices. His work is concentrated on comparing the magnitude of EPU shocks worldwide, and it was apparent that EPU in the U.S. has a greater magnitude compared to the one exerted by a Euro area-specific uncertainty shock. Thus, U.S. EPU could directly influence GCC countries because the U.S. is the main exporter to the GCC. Imports of goods in Saudi Arabia, Kuwait, and the UAE from the U.S. accounted for 11.8%, 9.1%, and 7% of the total imports, respectively (WITS, 2019). Hence, it is plausible to assume that U.S. EPU will have a direct influence on GCC economic indicators.

In this paper, I scrutinize the effect of U.S. EPU on six GCC countries starting from 1980, using an error correction model (ECM). For the top decile, income inequalities in GCC are moved partially by U.S. EPU, and to a lesser extent, by other macroeconomic aggregates, like inflation, both in the short and long run. However, for the top one percent earners, U.S. EPU does not have effects on their income distributions, both in the short and long run. As for the top one percent, the value of stocks traded can widen the income distribution. The results also show that other institutional variables, like education, play an important role in shrinking the income gap for the top one percent in the long run. Surprisingly, both in the long and short run, gross domestic product (GDP) growth does not have any effects on income distribution in GCC countries.

Findings of this paper have an important implication for the economics of inequality in GCC. They imply that writers should pay more attention to uncertainty, and particularly EPU. The rest of the paper follows this organization: first, I describe the methods used to measure income inequality, economic uncertainty. Second, I present my empirical model. Third, I show the results. Fourth, I discuss the economic effects of the results and conclude.

Measuring Income Inequality

Given the lack of income inequality data in GCC countries, the best alternative is to use the income inequality data housed at (WID). This indicator expresses how much each group controls out of the total income share within a country (deciles). Fortunately, WID provided income inequality data for all GCC since 1980.

Because household surveys do not usually observe inequality dynamics, WID observes it by combining some data sources, like fiscal, and survey data, along with national accounts, and wealth rankings, to capture the dynamics of various incomes from the least percentile to the highest.

An alternative to the concept of GDP, to compare the level of economic welfare across nations, WID uses the concept of national income (NI) for two reasons.² First, to make accurate welfare comparisons, the usage of NI is practical because it ignores: capital stock depreciation as it does not represent an actual income to individuals, and the share of domestic GDP paid to foreigners, capital owners. Second, GDP deals exclusively with aggregates and averages. It does not provide any useful information about how different income percentiles accumulate wealth after economic booms/expansions/.

World Income Database developed a methodology based on the concept of Distributional National Accounts (DINA).³ Its main goal is to explain the dynamics of the distribution of NI and wealth using notions of income and wealth that are plausible for comparison across different economies.⁴ Figure 1 suggests that, compared to the world, income inequality in Qatar and the UAE has risen substantially. However, Saudi Arabia maintained a lower than the world average income inequality for a substantial amount of time. Bahrain, Oman, and Kuwait had a steady level of income inequality that remained higher than the world average.

Figure 1.

Income distribution in GCC countries and the world (1980–2019) Source: World Inequality Database, 2019

Measuring Economic Uncertainty

Given the contradictions in measuring uncertainty, I use the U.S. EPU index recently constructed by (Baker et al., 2016) to model the indicators of economic policy uncertainty. U.S. EPU index was centered on the following factors: newspaper reporting, upcoming expirations of federal tax code requirements, and variation in potential economic forecasts.

Newspaper coverage is indexed based on the search outcomes of the major 10 newspapers.⁵ Construction of the index required monthly data of the given newspapers for vocabularies related to economic and strategy uncertainty. The aim of the search is to target specific terminologies like “uncertainty” or “uncertain,” and “economic” or “economy” and one or more of the following words: “congress,” “legislation,” “white house,” “regulation,” “federal reserve,” or “deficit.” Next, the quantity of article pieces that contain issues related to uncertainty is divided by the overall amount of article items in the respected paper. Afterward, the value in every month is normalized to obtain a multi-paper index.

Including tax code provisions is reasonable because they could cause uncertainty for individuals and firms. Typically, congress extends those provisions just before they expire “last minute extension,” which will trigger instability in the tax code. Tax code expiration data depends on information provided by the Congressional Budget Office (CBO). The procedure includes compiling lists of short-term federal tax provisions.

Economic expectation variation indicator depends on the Federal Reserve Bank of Philadelphia’s Evaluation of Expert Analysts. It measures the scattering for three predicted indicators that get affected by fiscal policies: inflation, local government expenditure, and federal government expenditure. Choosing these variables is meaningful as they are directly impacted by fiscal/monetary policies.

Finally, an overall index of EPU is assembled by standardizing each element by its standard deviation. Then, the average value of the element is computed by using different weights of the measures.⁶ Figure 2 provides U.S. EPU levels since 1985. It is apparent that U.S. EPU has reacted to key events in the GCC region.⁷

Figure 2.

U.S. economic policy Uncertainty Index (1985–2019).

U.S. EPU and Income Inequality: An Empirical Evaluation of 6 GCC Economies

I study how U.S. EPU impact income inequality via a panel analysis of 6 GCC economies between 1980 and 2019.⁸ Because EPU (along with other socioeconomic indicators) could have serious impacts on income distribution, Both in the short and long run, I use an ECM as an empirical estimator. Error correction model assumes a co-integrated relationship between the dependent and independent variables. In other words, stationarity appear after taking the first difference of the given variables. Thus, they exhibit a relationship (equilibrium) in the long run that could be disturbed and thus deviate in the short run (Box-Steffensmeier, Freeman, Hitt, & Pevehouse, 2014; Durr, 1992; Keele & De Boef, 2004).

The standard ECM is originated from a first order autoregressive (AR(1)) and distributed lag process, and the dependents are the first difference and need to be stationary.⁹ Hence, I include the independents two times in the model: first as a first difference, then as a lagged level.¹⁰ The standard (ECM) could be presented as follows

Δ Y_{i, t} = β_{0} + β_{1} Y_{i, t - 1} + β_{2} \sum Δ X_{i, t} + β_{3} \sum X_{i, t - 1} + β_{4} \sum C E_{i} + β_{5} \sum T E_{t} + ε

(1)

Δ Y_{i, t}

is the first difference of the proxy of income inequality for country i in year t.¹¹

Y_{i, t - 1}

, the lagged level of income inequality in country i in year t-1, is the error correction.¹² A negative significant effect of the lagged level of the independent variable is necessary, else, there is not a significant alteration back to equilibrium in the long run.

The vector of the first difference of the independents is $\sum Δ X_{i, t}$ for country i at time t. $β_{2}$ is the instant outcome of a variation in an explanatory variable on the change in income inequality. The vector of the lagged levels of the independents is $\sum X_{i, t - 1}$ for country/economy i at time t-1. They represent the long-run impacts of the explanatories on changes in inequality. It is important to note that the long-run outcome of the independents is not solely reliant on their parameter coefficients $(β_{3})$ but too the error correction, in which $(β_{1})$ suggests the degree of the change in the long run. Thus, the overall long-run impact of the independents can be computed by the ratio of $β_{3}$ to $β_{1}$ , negative ratio (Vlandas, 2018, p. 539). Since the long-run impact of the independents is calculated by the ratio of two coefficients rather than one, its standard error will follow, which denotes the significance of the variable in the long run.

Gross domestic product growth rate, inflation, returns on assets, unemployment, the proportion of elderly and human capital can influence the level of inequality in a country; therefore, I incorporate them in the model. To capture the return on capital, I incorporate the value of the stocks traded in a country to controls for asset income (Jorda, Knoll, Kuvshinov, Schularick, & Taylor, 2019). Data on the value of stocks traded are stemming from the World of Federation of Exchange Database (WFED, 2019). Data on GDP growth rate and inflation are stemming from (IMF, 2019). Data on the percentage of tertiary enrollment rate are stemming from the UNESCO institute of statistics (UNESCO, 2019). Data on unemployment are coming from International Labor Organization Database (ILOD, 2019). Data on the percentage of elderly people are coming from United Nation’s Population Division (UN, 2019).¹³

To control for omitted variables that may lead to an unexpected or suppressed growth in income inequality share that are invariant over time but vary across economies, I include a vector of country fixed effected resembled by $\sum C E_{i}$ . This method allows for more control of the national regulatory framework, the system of capitalism, and welfare regimes that influence income distribution share across different economies but do not differ much over time.

The vector that is represented by $\sum T E_{t}$ is an (n-1) time dummies that control for omitted time varying shocks. Hence, the years that led to sudden rise and fall of income across GCC countries will be accounted for.¹⁴ A likelihood ratio (LR) test¹⁵ imply the existence of heteroskedasticity within panels. Therefore, the country clustered standard errors are incorporated in both models. Finally, without the dependent variable, the independents are standardized so that a clearer image of comparison between the size of short- and long-run outcomes throughout all explanatory variables will be apparent. Thus, $(β'_{s})$ are understood as a “one standard deviation change in X widens the income inequality gap by a beta”.

Results

The findings of this paper are specified in Table 1. Model I investigate the effect of a one standard deviation variation in X on the top decile (p90) in GCC countries. In line with my theoretical expectation, U.S. EPU demonstrates a significant impact on income inequality in GCC both in the short and long run. Based on the parameters in Table 1, a one standard deviation increases in U.S. EPU means an instant rise in the income inequality for the top 10% earners (p90) in GCC countries. However, U.S. EPU has no significant effect, in both long run and short run, on income inequality for the top one percent earners (p99) in GCC countries.¹⁶

Table 1.

The short- and long-run determinants of movements in income inequality.

	I	II
Top earners (t-1)	−0.294** (−0.115)	−0.371** (0.13)
$Δ S t o c k s t r a d e d$	−0.00005 (0.000)	−0.0001 (0.000
$S t o c k s t r a d e d (t - 1)$	−0.00003 (0.000)	0.0002*** (0.000)
$Δ U S E P U$	0.0003* (0.000)	−0.0001 (0.000)
$U S E P U (t - 1)$	0.001** (0.000)	0.0002 (0.000)
$Δ G D P g r o w t h$	−0.0004 (0.000)	0.0004 (0.000)
$G D P g r o w t h (t - 1)$	−0.0004 (0.000)	0.0001 (0.000)
$Δ U n e m p l o y m e n t$	0.0002 (0.001)	0.001 (0.001)
$U n e m p l o y m e n t (t - 1)$	0.001 (0.001)	0.0003 (0.0004)
$Δ I n f l a t i o n$	−0.0003*** (0.000)	−0.0002 (0.000)
$I n f l a t i o n (t - 1)$	−0.001*** (0.000)	−0.0005*** (0.000)
$Δ E n r o l l m e n t$	−0.0002 (0.000)	−0.001 (0.000)
$E n r o l l m e n t (t - 1)$	−0.0006 (0.001)	−0.002* (0.001)
$Δ E l d e r l y$	−0.001 (0.01)	−0.007 (0.01)
$E l d e r l y (t - 1)$	0.0002 (0.001)	0.001 (0.001)

Unsurprisingly, inflation demonstrates a significant impact on income inequality in GCC countries both in the short and long run, for both the top decile and the top one percent earners. Based on the parameters in Table 1, a one standard deviation rise in the price level causes an instant decline in income distribution for the top decile, and the top one percent earners. This contradicts the findings of (Fuller, Johnston, & Regan, 2020) that inflation shapes the distribution of higher wealth inequalities.

In contrast to the top decile, in the long run, the value of stocks traded in the stock market has a significant impact on income distribution inequality for the top one percent. As shown in Table 1, a one standard deviation rise in the value of stocks traded leads to an immediate rise in the income inequality proportion. This challenges what (Huber, Huo, & Stephens, 2017; Piketty, 2014) claimed that asset values only form the dynamics of wealth inequalities in the short run (i.e., booms/busts). Somehow surprisingly, the education level does not behave as (Fuller, et al., 2020) would predict. The Rising tertiary enrollment rate has no long-run nor short-run effects on income inequality for the top decile. However, tertiary enrollment would decrease the income inequality rate in the long run for the top one percent earners. Finally, GDP growth has a negligible impact on the income distribution level for both the top decile and the top one percent both in the short and the long run. This contradicts what Piketty anticipated that real GDP (income) growth decreases the wealth concentration ratio.

The Dependent is the first difference of the income inequality rate. Independents are standardized; estimator used was an ECM for 6 GCC countries from 1980 to 2019. (N-1) time dummies, (n-1) country dummies, and the constant not shown. p-value *, **, *** indicate significance at 90%, 95% and 99% confidence level.

Discussion and Conclusion

The findings have major contributions for the literature on inequality and EPU in GCC countries. First, they imply that principles on EPU are well equipped to explain the dynamics of wage inequality. It is not GDP growth rates, nor the ratio of elderly, nor the unemployment rate that directly moves income inequality in GCC. Instead, it is U.S. EPU. For the first glance, the top income earners in GCC seem to be benefiting from high uncertainty in the U.S.; however, this may not be the case. In other words, it might be the lower 90% income earners who are suffering from rising uncertainty levels in the U.S. When uncertainty levels rise in the U.S., this might lead the lower 90% earners (weaker bound) to lose their jobs or to face a wage reduction, while the top decile’s income does not change, or even decreases but at a lesser magnitude than the reduction the lower bound faced.

Second, the results also indicate that asset prices impact income distribution in GCC countries. While stock prices do not seem to affect the top decile, it directly impacts the top one percent in the long run. This finding ignites questioning the nature of stock owners in GCC. In other words, if stocks “substantial” owners are typically the top one percent, then it would be logical that changes in stock prices lead to widening the income gap on the upper level as the lower 99% do not seem to worry about carrying stocks in their portfolios. Usually, earnings from bonds/stocks or other type of assets outgrow NI, thus wealth/income will become more tilted toward owners of stocks, bonds, or other assets rather than labor power (Piketty & Saez, 2003).

The ongoing pandemic of COVID-19 created worldwide uncertainty concerns that of question: “How could global uncertainty impact the labor market in GCC and other countries? “This study showed that uncertainty in the U.S. can play an important role in explaining the income distribution in GCC. Moreover, the topic of U.S. uncertainty impact on GCC region needs further investigations.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Omar Alenezy

Notes

References

Alesina

Rodrick

(1994). Distributive politics and economic growth. Quarterly Journal of Economics, 109(2), 465–490. https://doi.org/10.2307/2118470.

Baker

S. R.

Bloom

Davis

S. J.

(2016). Measuring economic policy uncertainty. The Quarterly Journal of Economics, 131(4), 1593–1636. https://doi.org/10.1093/qje/qjw024.

Banerjee

A. V.

Duflo

(2003). Inequality and growth: What can the data say? Journal of Economic Growth, 8(3), 267–299. https://doi.org/10.1023/a:1026205114860.

Beck

Demirgüç-Kunt

Levine

(2007). Finance, inequality and the poor. Journal of Economic Growth, 12(1), 27–49. https://doi.org/10.1007/s10887-007-9010-6.

Box-Steffensmeier

Freeman

Hitt

Pevehouse

(2014). ‘Cointegration and error correction models’, time series Analysis for the social sciences (pp. 150–172). Cambridge: Cambridge University Press.

Colciago

Samarina

de Haan

(2019). Central bank policies and income and wealth inequality: A survey. Journal of Economic Surveys, 33(4), 1199–1231. https://doi.org/10.1111/joes.12314.

Colombo

(2013). Economic policy uncertainty in the US: Does it matter for the Euro area? Economics Letters, 121(1), 39–42. https://doi.org/10.1016/j.econlet.2013.06.024.

De Haan

Sturm

J. E.

(2017). Finance and income inequality: A review and new evidence. European Journal of Political Economy, 50(2017), 171–195. https://doi.org/10.1016/j.ejpoleco.2017.04.007.

Demirgüç-Kunt

Levine

(2009). Finance and inequality: Theory and evidence. Annu. Rev. Financ. Econ, 1(1), 287–318. https://doi.org/10.1146/annurev.financial.050808.114334.

10.

Durr

R.H.

(1992). An essay on cointegration and error correction models. Political Analysis, 4(1992), 185–228. https://doi.org/10.1093/pan/4.1.185.

11.

Fawaz

Rahnamamoghadam

Valcarcel

(2012). Fluctuations, uncertainty and income inequality in developing countries. Eastern Economic Journal, 38(4), 495–511. https://doi.org/10.1057/eej.2011.32.

12.

Fawaz

Rahnama

Valcarcel

V. J.

(2014). A refinement of the relationship between economic growth and income inequality. Applied Economics, 46(27), 3351–3361. https://doi.org/10.1080/00036846.2014.929624.

13.

Federal Open Market Committee . (2009). Minutes of the December 2009 meeting. http://www.federalreserve.gov/monetarypolicy/fomcminutes20091216.htm.

14.

Fuller

G. W.

Johnston

Regan

(2020). Housing prices and wealth inequality in Western Europe. West European Politics, 43(2), 297–320. https://doi.org/10.1080/01402382.2018.1561054.

15.

Huber

Huo

Stephens

J. D.

(2017). ‘Power, policy, and top income shares’, socio-economic review. https://doi.org/10.1093/ser/mwx027.

16.

International Labour Organization . (2021). ILOSTAT database. Data retrieved on January 29.

17.

International Monetary Fund . (2013). World economic outlook: Hopes, realities, risks,” April. IMF Press.

18.

International Monetary Fund . (2019). https://www.imf.org/en/Home (Accessed June 2021).

19.

Jordà

Ò.

Knoll

Kuvshinov

Schularick

Taylor

A. M.

(2019). The rate of return on everything, 1870–2015. The Quarterly Journal of Economics, 134(3), 1225–1298. https://doi.org/10.1093/qje/qjz012.

20.

Keele

De Boef

(2004). ‘Not just for Co-integration: Error correction models with stationary data’, working paper. Nuffield College: Department of Political Science and International RelationsOxford University. https://www.nuffield.ox.ac.uk/politics/papers/2005/Keele%20DeBoef%20ECM%20041213.pdf.

21.

Piketty

(2014). Capital in the twenty-first century. Harvard: Harvard University Press.

22.

Piketty

Saez

(2003). Income inequality in the United States, 1913–1998. Quarterly Journal of Economics, 118(1), 1–41. https://doi.org/10.1162/00335530360535135.

23.

Samarina

Nguyen

A. D.

(2019). Does monetary policy affect income inequality in the euro area?

24.

UNESCO Institute for Statistics . (2020). Data as of September. http://uis.unesco.org/.

25.

Vlandas

(2018). Grey power and the economy: Aging and inflation across advanced economies. Comparative Political Studies, 51(4), 514–552. https://doi.org/10.1177/0010414017710261.

26.

World Federation of Exchange Database . (2019). https://www.world-exchanges.org/(accessed June, 2021).

27.

World Inequality Database . (2019). http://wid.world/ (accessed May–June, 2021).

28.

World Integrated Trade Solutions . (2019). https://wits.worldbank.org/ (accessed May-June 2021).