Nominal and discretionary household income convergence: The effect of a crisis in a small open economy

Highlights

• We study crisis induced changes in regional household income convergence.

• We design easy to apply and interpret conditional convergence test.

• Our test is applied to 43,188 households in a small open economy of Slovakia.

• While nominal income converged only before, discretionary has not converged before nor during the crisis.

• Results provide evidence that income and consumption need to be evaluated jointly.

Abstract

Regional economic imbalances have considerable negative effects on society that might be amplified during an economic recession. We study changes in the cross-regional household income imbalance in the small open economy of Slovakia before and during an economic recession. Apart from the usual nominal income, we study discretionary income, which accounts for consumption of necessity goods. Our analysis uses household budget survey data from January 2004 until December 2012 with 43,188 overall observations. Our main results are based on the $\sigma$-convergence framework, and we observe that only nominal regional household income converged before the economic recession. However, during the crisis, neither nominal nor discretionary household income converged. The $\beta$-convergence framework shows that only discretionary income converged during both periods. These results support the hypothesis of Stiglitz et al. (2009) that income and consumption need to be evaluated jointly to assess (regional) economic imbalances and inform the design of appropriate policies.

Introduction

As regional economic imbalances might have severe economic, social and political consequences, there is a rich field of research on regional disparities and convergence aimed at yielding relevant policy recommendations. Conventional approaches to the estimation and analysis of regional economic disparities are based on the neoclassical growth theory of Solow (1956) and Swan (1956), as well as the seminal studies of Barro et al. (1991) and Barro and Sala-i Martin (1992). The main stream of literature is based on the concept of income convergence between different regions (countries); see, e.g., Kočenda et al. (2006), Ertur and Koch (2006), Niebuhr and Schlitte (2009), Cavenaile and Dubois (2011), Crespo Cuaresma and Feldkircher (2013), Matkowski et al. (2016), Iammarino et al. (2019), Ezcurra (2019), Cutrini (2019) or Cieślik and Wciślik (2020). The findings on convergence tend to differ depending on the methodology used and the geographic region and period covered. Moreover, the impact of economic crises on regional economic inequalities is not well understood.

The most widely employed measure of economic activity within studies of regional disparities is gross domestic product (GDP) per capita, even though it was not conceived as a measure of welfare or well-being. Instead of focusing on regional GDP, which has limitations in this respect (e.g., it ignores the income distribution, nonmonetary production and environmental costs), Stiglitz (2016) suggests focusing on how the economy is performing for a typical citizen. Stiglitz et al. (2009) argue that the household perspective provides different, policy-relevant information, as in many open and export-oriented economies (including Slovakia), real disposable income does not track with GDP growth particularly well. Aitken (2019) states that household income diverges from per capita GDP growth in a number of OECD countries, including the US, and thus that reliance on the GDP measure increases the risk of inappropriate policy decision-making. This line of research has been followed by, e.g., Goedemé and Collado (2016), who suggest a different measure of social cohesion – equivalized disposable household income expressed in purchasing power terms.

In this study, our main contribution lies in using an alternative approach, most closely related to those of Beech et al. (2014) and Strout et al. (2019), to measurement of economic well-being. The idea is that a given household’s economic well-being depends not on the amount of money earned but on the amount of money left after all necessary expenses are covered. We refer to this income as discretionary income. As discretionary income captures not only income but also consumption, it is reasonable to assume that nominal and discretionary household income convergence would differ. As regional consumer price indices are rarely available or are of poor quality, which is also the case in our empirical context, discretionary income may be used as a viable complement or alternative to existing measures, as it overcomes the limitations of GDP per capita and nominal income by accounting for heterogeneous regional price levels and different household preferences.

In particular, in the case of an economic crisis, households may be affected by induced changes in their income (e.g., a higher portion of redistribution in total income or a decline in total income), consumption (e.g., a decrease in the consumption of nonessential goods and services, postponement of durable good purchases) or both. The effects likely vary depending on each household’s and region’s characteristics. Moreover, depending on its sources, a crisis might influence industries, regions and households differently. For example, the crisis that started in 2008 originated in the financial sector and affected investment activities (e.g., Burger et al., 2017). On the other hand, the 2020 economic crisis induced by the COVID-19 pandemic is likely to have much more severe consequences for specific industries, such as tourism and transportation, and consequently affect different regions and households in a considerably more heterogeneous way.¹ Given the complexity of economic crises, one would expect regional income convergence to be disrupted by such forces.

Existing literature on the effects of economic crises on income inequality seems to focus on cross-country comparisons. Raitano (2016) and Caminada et al. (2019) find increasing income inequality in postcrisis periods, although redistribution policies seem to reduce economic imbalances (see, e.g., Caminada et al., 2019). Vacas-Soriano and Fernández-Macías (2018) suggest that the income convergence process among EU countries halted during the 2009–2012 crisis period. Bolea et al. (2018) point to a significant break point in the process of European convergence with the arrival of the economic crisis in 2008. Several recent studies suggest that the latest crisis might increase income inequality. For example, Mussida et al. (2016) provide evidence on the exacerbating effects of the economic crisis on income inequality among individuals within and between Italian regions.

Our study contributes to this strand of the literature. We study how the economic crisis disrupted regional economic imbalances at the household level across regions of Slovakia, a small but, given its size, heterogeneous economy. We estimate and compare regional income convergence within two periods: the precrisis period (2004–2008) and the crisis period (2009–2012). In doing so, we make two novel contributions to the literature. First, we design a convergence test that compares the conditional predicted regional income variation between two periods of time; thus, the test belongs to a class of $\sigma$-convergence tests.² Second, we not only study convergence based on household-level nominal incomes but also extend the analysis to discretionary income, defined as nominal household income less consumption on short-term necessities (e.g., groceries, apparel, energy, rent, and financial service costs). While nominal income is commonly employed in the regional income inequality literature, it does not reflect the fact that in more developed regions, price levels might also be higher (e.g., property costs may be higher in more densely populated and larger cities). The use of nominal income in an income growth study framework might therefore lead to an overstatement of regional imbalances. As household characteristics drive consumption preferences (van Leeuwen et al., 2020), we account for variations in household characteristics that are captured with monthly unbalanced budget survey household data over the period from 2004 to 2012.

Our results suggest that the distinction between nominal and discretionary income is relevant, as the two measures lead to different conclusions. Specifically, we find that in the precrisis period, regional nominal income converged ($\sigma$-convergence), but discretionary income did not. During the crisis period, regional nominal or discretionary income converged. The results on the $\beta$-convergence show that nominal income did not converge, while discretionary income did converge before and during the crisis period. It therefore appears that the economic crisis had a heterogeneous impact on different measures of economic well-being.

The remainder of the paper is organized as follows. In the next section, we discuss income and consumption as measures of well-being and the corresponding weaknesses that lead us to use discretionary income. The following section focuses on the data and methodology used for estimation of the effects of the crisis on regional inequality development. Our results together with the limitations of our approach are discussed in the subsequent sections, and a summary with possible implications is elaborated in the conclusion.