“Economic growth and housing spending within social protection: Correlation and causal study”

The purpose of this study is to ground the causality, its character, and power between economic growth and housing spending within government social protection to strengthen poverty reduction. The study is conducted on a sample of 27 EU countries for 2012–2021 according to the following indicators: government expenditure on housing within social protection, government expenditure on housing development, and GDP per capita growth. Correlation analysis revealed the relationship between the variables. The paper employed time lags with the most significance based on Pearson and Spearman correlation coefficients depending on normal data (Shapiro-Wilk test). The causal analysis determined which of the studied indicators is the cause and consequence of established dependence based on the results of the Granger test. Calculations are made in STATA software. It is confirmed that government expenditure on housing within social protection influences GDP per capita growth in 14 countries (the highest impact is in Greece and Spain) and government expenditure on housing development – in 17 countries (the biggest influence is in Romania, the Slovak Republic, Ireland, and Lithuania). It is also emphasized that government expenditure on housing development influences economic growth more significantly and with higher strength than government expenditure on housing within social protection. The obtained results can be useful in further research and government decision-making in social and economic policy, particularly regarding the expediency of increasing government spending for affordable housing and its development according to social protection programs, poverty reduction, and inclusive economic growth.


INTRODUCTION
The problem of affordable and social housing has always existed, but in global challenges, such as wars, pandemics, or economic crises, it becomes increasingly acute. Kostenko et al. (2022) grounded the relevance of social resilience during the Covid-19 pandemic. Pakhnenko et al. (2022) distinguished the role and peculiarities of financial aid to communities considering the Covid-19 pandemic. Zhang et al. (2022) emphasized changes in household expenditure and migrants' remittances caused by Covid-19 consequences. Voznyak et al. (2023aVoznyak et al. ( , 2023b) paid attention to the migration processes, especially during the war, and accordingly, the increasing necessity of social housing based on survey results that showed a lack of affordable housing in the community (68.6% of responses). Alekseyenko et al. (2021) also substantiated the need for affordable housing for internally displaced persons.
Besides, the urgent need to improve housing affordability for youth and young families is driven by the independence from parental support for financial ability and housing (Kutlák, 2021) and by housing overcrowding, overpopulation, and urbanization . The health situation is closely connected with housing affordability and conditions Ovcharova & Grabowska, 2022;Sheliemina, 2023). Thus, changes in public policy and strategies for post-war and other crises overcoming are significant, particularly in social interactions . The provision, management, and financing of social and affordable housing is no exception.
To improve the outlined situation, a significant role is assigned to the government, which makes decisions on changes in social, housing, and communal policies, as well as budget policy, revising the structure of budget expenditures, one of the items of which is the financing of housing development, housing construction and subsidizing housing for the poor or unprotected population groups. Therefore, the issue of social protection and housing financial aid for poverty reduction and economic growth is relevant and timely.

LITERATURE REVIEW
Social protection is positioned as a factor of sustainable development (Tu et al., 2023). Reducing poverty due to social improvement is one of the sustainable development goals for today (Awojobi, 2022; Makole et al., 2022;Salisu, 2022 Marica and Piras (2018) conducted a bibliometric review and investigated some empirical as-pects of the impact of public spending on economic growth. Poku et al. (2022) analyzed the dependence between public expenditure and economic growth in Ghana based on the data for 1970-2016 and the ARDL econometric estimate. They concluded that government spending is in a positive relationship with economic growth in the short period. At the same time, population growth has a negative relationship with economic growth. Musaba et al. (2013) investigated the impact of public spending on economic growth in the case of Malawi using time series from 1980 to 2007. However, the investigated list of expenses did not include an assessment of expenses for social security in general and financing of social housing in particular. Li (2018) and Libertun de Duren and Osorio (2020) studied the influence of local budget spending on housing prices and housing deficits. Tahat (2023) grounded the dependence between housing demand and macroeconomic factors. Egan and Bergin (2023) examined the effect of government spending on the housing supply in Ireland and the effects of the government's impact on the real estate market across the economy and by sector. Doling et al. (2013) studied the connection between housing, financing, and economic development. The benefits of financing housing have been found to occur indirectly through improved health, and multiplier effects of conditions of employment and production. A symbiotic relationship between the financing of housing construction and the development of the financial sector was also revealed. Kumar (2021) measured the impact of the housing program, particularly the subsidized housing lottery in Mumbai (India) on the population's welfare. Kunovac and Zilic (2022) investigated the effect of subsidizing housing loans in Croatia. They argued that the subsidy disrupted the normal intra-annual dynamics of housing transactions, and house prices rose immediately after the subsidies were introduced. Molidya and Fanggidae (2020) researched the optimization of state subsidies in the housing sector based on the experience of Kupang during the Covid-19 pandemic.
However, the issue of social protection and housing financial aid by the government is not investigated enough and requires further development. Government housing spending is vital for improving the living conditions of citizens and reducing poverty in general. Funding social and affordable housing from budget expenditures is one of the most widespread sources of strengthening housing affordability. At the same time, financing these items of expenditure occurs along with financing a significant list of other state functions. Therefore, it is appropriate to investigate whether there is a relationship between these expenditures and the country's economic growth.
The purpose of this study is to ground the causality, its character, and power between economic growth and housing spending within government social protection to strengthen poverty reduction.

METHODS
The causality between housing financial aid according to government social protection and economic growth was determined based on the sample of 27 countries -current members of the European Union (European Union, n.d.).
The period of investigation was 2012-2021, i.e., the last ten years for which Eurostat and the World Bank have published statistical data according to the following indicators: • GE_H -government expenditure on housing (in the spending block of social protection) as a percentage of total general government expenditure (Eurostat, 2023); • GE_HD -government expenditure on housing development (in the spending block of housing and community amenities) as a percentage of total general government expenditure (Eurostat, 2023); • GDP -GDP per capita growth as an annual percentage (World Bank, n.d.).
To confirm or reject the relationships between housing financial aid according to government social protection and economic growth (in particular, between government expenditure on housing in the spending block of social protection and GDP per capita growth and between government expenditure on housing development in the spending block of housing and community amenities and GDP per capita growth) and identify their nature (direction), strength and time lags with its most significance, the study used correlation analysis based on calculating Pearson  The grounding of which of studied indicators is the cause and which is the consequence in established dependences was conducted based on causal analysis -the Granger test (Granger, 1969

RESULTS
Government expenditure on housing in the spending block of social protection occupies a small share of the total volume of state expenses. At the same time, they differ significantly within the sample of EU countries ( Figure 1).
The highest level of housing financial aid, according to government social protection, is in Ireland, France, Denmark, Finland, and the Netherlands. This share is equal to zero in such EU countries as Romania, the Slovak Republic, Slovenia, and Spain. So, the gap is significant even within the EU. Next, Figure 2 shows the comparative analysis of government expenditure on housing development in the spending block of housing and community amenities.
The leading EU countries on this issue are Bulgaria, Romania, Luxemburg, France, Latvia, Ireland, and the Republic of Cyprus. In Greece and Finland, this item of expenditure is equal to zero. However, in the case of government expenditure on housing development compared with government expenditure on housing as social protection, the gap between countries is not so noticeable and significant. Besides it, some countries, such as Ireland, have a high level of government spending. In turn, Romania is a leader in government expenditure on housing development, but at the end of the list in the case of government expenditure on housing as social protection. The same reverse situation takes place in Finland. This is due to differences in the state social policy, social protection programs, in particular, regarding the financing of social and affordable housing, as well as various mechanisms, tools, forms, methods, and sources of its provision. All above is very important to reduce poverty and housing overcrowding.
Taking into account the given differences in the priorities of state financing of social housing -either within the limits of social protection (housing subsidies, etc.) or within the limits of the development of housing construction, it is vital to investigate the causal relationships between housing financial aid according to government social protection and economic growth to strengthen householders welfare and poverty reduction.
Firstly, the input data of government expenditure on housing in the spending block of social protection (GE_H) (   If Prob > z is less than 0.5, the data are not subject to the law of normal distribution. It is important for choosing a method of correction analysis at the next research stage: -Pearson method is applied when the data are normal (Prob > z is equal or more than 0.5); -Spearman method is applied when the data are not normal (Prob > z is less than 0.5).
Secondly, correlation analysis was conducted to confirm or reject the existence of a relationship between housing financial aid according to government social protection and economic growth for each country from the sample and to identify the nature of the relationship (direct or reversed) and strength (low, middle, or high). Time lags were also applied to determine when the relationship has the highest significance level. For these calculations, STATA instruments were used ( Table 2).
The dependence between government expenditure on housing in the spending block of social protection and GDP per capita growth was con-   The causal study was made to find which of the above indicators (result and factor) is the cause and which is the consequence. For this purpose, the Granger causality test and the Vector Autoregression (VAR) were applied using STATA software. It contributes to determining which variable is a determining factor for another one or, in other words, identifying the direction of the causality link (Table 3).  -the input data are unchanged during the studied period or are equal to zero; * -the investigated factor is not the cause of the resulting indicator; "-" -causality has not been established.

Country name Between GE_H and GDP Between GE_HD and GDP
For example, for the first pair of variables (the 1 st column), if the p-value is less or equal to 0.05, the hypothesis that GE_H is a cause of GDP is confirmed. If the p-value is more than 0.05, the investigated factor is not the cause of the resulting indicator, and causality was not established. Therefore, the direction of causality is from GE_H (government expenditure on housing in the spending block of social protection) to GDP (GDP per capita growth). The same rule was used to estimate the results presented in the other three columns for two pairs of investigated indicators. In some countries, there is causality in both directions; it is a bidirectional causality. Table 4 shows the generalized and visualized results of the casual study.

Causality between GE_HD and GDP
Austria -

GE_H → GDP GE_HD → GDP
Note: "-"-causality is not confirmed or cannot be established at a given time interval in a specific country.
Therefore, government expenditure on housing in the spending block of social protection is a cause of GDP per capita growth in 6 EU countries. The reverse causality (GDP per capita growth is a cause of government expenditure on housing as social protection) takes place in 5 countries. Besides, there is bidirectional causality in 8 countries. In the end, government expenditure on housing in the spending block of social protection influences GDP per capita growth in 14 EU countries.
In turn, government expenditure on housing development in the spending block of housing and community amenities is a cause of GDP per capita growth in 6 EU countries. The reverse unidirectional causality (GDP per capita growth is a cause of government expenditure on housing development in the spending block of housing and community amenities) exists in 7 countries. In addition, there is bidirectional causality in 11 countries. So, government expenditure on housing development in the spending block of housing and community amenities influences GDP per capita growth in 17 EU countries. Ianchuk et al. (2021) confirmed a hypothesis about the impact of owners' number of housing with loans on economic growth and the number of tenants, rent social (free) housing, or housing with reduced prices on economic growth. The conclusion showed that government should pay more attention to affordable lending housing strategies over programs to reduce rental housing. However, this study examines completely different mechanisms of financing social housing and their impact on economic growth: government expenditure on housing development and government expenditure on housing in the spending block of social protection.

DISCUSSION
Chugunov and Nasibova (2021) also studied the issue of government financing of social protection on the evidence of European countries. They concluded the influence on indicators of socio-economic development, paying attention to the inconsistency in the amount of financing social protection expenditures. At the same time, the subject of their study was social payments and contribu-tions, the structure of social protection, not including the issue of financing social housing and related problems.
Hasnul (2015)  This study has some limitations. Firstly, it has a limited sample of the study countries -EU members. Secondly, the research period is limited toothe last ten years, for which the investigated indicators were published. Therefore, in further studies, it is advisable to expand the sample of countries, paying attention to developing countries, where the problem of poverty, lack of healthy living conditions, and housing overcrowding is even more acute. Accordingly, the increase of state expenditures under various budget items on social and affordable housing can significantly impact economic growth more than in developed countries.

CONCLUSION
This study grounded the causality, character, and power between economic growth and housing spending within government social protection to strengthen poverty reduction. Correlation analysis confirmed the relationship between government expenditure on housing as social protection and GDP per capita growth in 22 from 27 EU countries, including 14 countries with positive dependence, mostly middle strength and 1-year lag (or without time lag). In turn, the interconnection between government expenditure on housing development in the spending block of housing and community amenities and GDP per capita growth was determined in 26 from 27 sample countries, which involve 17 ones with positive direction, high (8 countries) and middle (9 countries) strength and 1-2-year lag. Using the results of Granger causality testing, government expenditure on housing as social protection influences GDP per capita growth in 14 EU countries, and government expenditure on housing development in the spending block of housing and community amenities is a cause of GDP per capita growth in 17 EU countries. Greece and Spain have the highest impact of government expenditure on housing as social protection on GDP per capita growth. The most significant influence of government expenditure on housing development on economic growth is in Romania, the Slovak Republic, Ireland, and Lithuania.
Comparing the economic growth of these two indicators of financing social and affordable housing from government spending, it is noted that government expenditure on housing development influences GDP per capita growth more significantly and with a higher level of power than government expenditure on housing as social protection. Therefore, to strengthen economic growth, improve the well-being of households and overcome poverty, it is more appropriate for the governments of EU countries to invest in the development of the housing stock and housing construction than to pay housing subsidies within the framework of social protection programs. The obtained results can be useful in further research and government decision-making in social and economic policy, in particular, regarding the expediency of increasing government spending for affordable housing and its development according to social protection programs, poverty reduction, and inclusive economic growth.