The analyses included in this thesis contribute to the literature on fiscal policies and economic growth through different angles, where we mainly take advantage of i) the longer fiscal time series that largely also cover the post Great Recession period; ii) data of interest that involve a larger share of developing countries compared to most of the previous studies; iii) the availability of novel econometric techniques for macro panel data. This dissertation is made up of four chapters who address different, though related, phenomena. The empirical analyses are conducted at panel level within all the chapters. The first chapter examines the impact of the debt-to-GDP ratio on aggregate growth, productivity and capital accumulation. Estimates are conducted at annual and 5-years (overlapping and non-overlapping) levels within a dynamic framework. Compared to most of the previous studies, we use a panel dataset that is larger both in N and in T, with consequently benefits for the inference. The overlapping and non-overlapping growth episodes are computed in a new way that maximizes the number of observations. Lastly, we address the problem of missing values through forward orthogonal deviation (FOD) techniques. The issue of gaps in the sample represents a concern when datasets contain developing countries – as many time series are relatively short compared to those of advanced economy. As a result, such a problem may threaten the inference when estimation techniques are based on the employment of internal instruments. The second chapter investigates the short-run and long-run effects of public debt per worker on output per worker while accounting for country-heterogeneity, nonstationarity, endogeneity and cross-sectional dependence through novel econometric techniques. We derive an estimable equation from a Cobb-Douglas output function augmented with debt where the parameters are allowed to vary across countries. We consider both gross and net public debt. Moreover, we estimate asymmetric models to examine whether the processes of debt accumulation and relief follow a nonlinear path. The third chapter studies the long-run effects on the private sector of the current government expenditure within a dynamic heterogenous framework where errors are assumed to be cross-sectionally dependent. We consider both the aggregate government expenditure and its bipartition into productive and unproductive. Moreover, we also isolate the private investment effects of the mere reallocation of public resources towards each of the spending categories by holding constant the size of the government intervention. Finally, we build a government budget constraint (GBC) in order to examine whether the way government expenditure is financed matters for the long-run dynamics of private investments. The fourth chapter establishes a cointegrating relationship between private investments and the single components of government expenditure – categorized according to the socioeconomic objectives – where, as in the second and third chapters, the heterogenous effects of global shocks and local spillovers are accounted for. We alternatively consider the components of government expenditure both as share of GDP and as share of total government expenditure. This distinction matters because the effect of the single components as share of total government expenditure should be considered as the effect of a reallocation of resources towards a given spending component, since the size of government intervention does not adjust. In addition, also this chapter includes the GBC in the regressions in order to obtain estimates of the response of private investments to the various combinations of spending components and methods of financing.
Le analisi incluse in questa tesi contribuiscono alla letteratura sulle politiche fiscali e sulla crescita economica da diverse angolazioni, dove si sfruttano principalmente i) serie storiche fiscali più lunghe che coprono in gran parte anche il periodo successivo alla Grande Recessione; ii) dati di interesse che coinvolgono una quota maggiore di paesi in via di sviluppo rispetto alla maggior parte degli studi precedenti; iii) la disponibilità di nuove tecniche econometriche per dati macro-panel. Questa dissertazione si compone di quattro capitoli che affrontano fenomeni diversi, sebbene correlati. Le analisi empiriche sono condotte a livello di panel all'interno di tutti i capitoli. Il primo capitolo esamina l'impatto del rapporto debito/PIL sulla crescita aggregata, sulla produttività e sull'accumulazione di capitale. Le stime sono condotte a livello annuale e quinquennale (overlapping and non-overlapping) all'interno di un quadro dinamico. Rispetto alla maggior parte degli studi precedenti, in questa tesi viene utilizzato un set di dati panel più ampio sia in N che in T, con conseguenti vantaggi inferenziali. Gli episodi di crescita sovrapposti e non sovrapposti sono calcolati con un nuovo metodo che massimizza il numero di osservazioni. Infine, affrontiamo il problema dei missing values attraverso tecniche di deviazione ortogonale in avanti (FOD). La presenza di missing values può rappresentare un problema quando i dataset contengono paesi in via di sviluppo, poiché molte serie temporali sono relativamente brevi rispetto a quelle per le economie avanzate. Il secondo capitolo indaga gli effetti a breve e lungo termine del debito pubblico per lavoratore sull’output per lavoratore, tenendo conto dell'eterogeneità tra le unità, della non stazionarietà, dell'endogeneità e della cross-sectional dependence attraverso nuove tecniche econometriche. Deriviamo un'equazione stimabile da una funzione di produzione Cobb-Douglas estesa con il debito e in cui i parametri possono variare tra i paesi. Consideriamo il debito pubblico lordo e netto. Inoltre, stimiamo modelli asimmetrici per esaminare se i processi di accumulazione e riduzione del debito seguono un percorso non lineare. Il terzo capitolo studia gli effetti a lungo termine sul settore privato della spesa pubblica corrente all'interno di un quadro dinamico eterogeneo in cui si presume che gli errori dipendano trasversalmente. Consideriamo sia la spesa pubblica aggregata che la sua bipartizione in produttiva e improduttiva. Inoltre, isoliamo anche gli effetti dell'investimento privato della mera riallocazione delle risorse pubbliche verso ciascuna delle categorie di spesa mantenendo costante la spesa pubblica totale. Infine, costruiamo un vincolo di bilancio pubblico (GBC) per esaminare se il modo in cui la spesa pubblica è finanziata è importante per le dinamiche di lungo periodo degli investimenti privati. Il quarto capitolo stabilisce un rapporto di cointegrazione tra gli investimenti privati e le singole componenti della spesa pubblica – classificate secondo gli obiettivi socioeconomici – dove, come nel secondo e nel terzo capitolo, si tiene conto degli effetti eterogenei degli shock globali e delle ricadute locali. In alternativa, consideriamo le componenti della spesa pubblica sia come quota del PIL che come quota della spesa pubblica totale. Questa distinzione è importante perché l'effetto delle singole componenti sulla quota della spesa pubblica totale va considerato come l'effetto di una riallocazione delle risorse verso una determinata componente di spesa, in quanto la spesa pubblica totale non si aggiusta nel modello. Inoltre, anche questo capitolo include il GBC nelle regressioni al fine di ottenere stime della risposta degli investimenti privati alle diverse combinazioni di componenti di spesa e modalità di finanziamento.
ESSAYS ON FISCAL POLICIES AND ECONOMIC GROWTH / Carvelli, Gianni. - (2022 Jun 13).
ESSAYS ON FISCAL POLICIES AND ECONOMIC GROWTH
CARVELLI, Gianni
2022-06-13
Abstract
The analyses included in this thesis contribute to the literature on fiscal policies and economic growth through different angles, where we mainly take advantage of i) the longer fiscal time series that largely also cover the post Great Recession period; ii) data of interest that involve a larger share of developing countries compared to most of the previous studies; iii) the availability of novel econometric techniques for macro panel data. This dissertation is made up of four chapters who address different, though related, phenomena. The empirical analyses are conducted at panel level within all the chapters. The first chapter examines the impact of the debt-to-GDP ratio on aggregate growth, productivity and capital accumulation. Estimates are conducted at annual and 5-years (overlapping and non-overlapping) levels within a dynamic framework. Compared to most of the previous studies, we use a panel dataset that is larger both in N and in T, with consequently benefits for the inference. The overlapping and non-overlapping growth episodes are computed in a new way that maximizes the number of observations. Lastly, we address the problem of missing values through forward orthogonal deviation (FOD) techniques. The issue of gaps in the sample represents a concern when datasets contain developing countries – as many time series are relatively short compared to those of advanced economy. As a result, such a problem may threaten the inference when estimation techniques are based on the employment of internal instruments. The second chapter investigates the short-run and long-run effects of public debt per worker on output per worker while accounting for country-heterogeneity, nonstationarity, endogeneity and cross-sectional dependence through novel econometric techniques. We derive an estimable equation from a Cobb-Douglas output function augmented with debt where the parameters are allowed to vary across countries. We consider both gross and net public debt. Moreover, we estimate asymmetric models to examine whether the processes of debt accumulation and relief follow a nonlinear path. The third chapter studies the long-run effects on the private sector of the current government expenditure within a dynamic heterogenous framework where errors are assumed to be cross-sectionally dependent. We consider both the aggregate government expenditure and its bipartition into productive and unproductive. Moreover, we also isolate the private investment effects of the mere reallocation of public resources towards each of the spending categories by holding constant the size of the government intervention. Finally, we build a government budget constraint (GBC) in order to examine whether the way government expenditure is financed matters for the long-run dynamics of private investments. The fourth chapter establishes a cointegrating relationship between private investments and the single components of government expenditure – categorized according to the socioeconomic objectives – where, as in the second and third chapters, the heterogenous effects of global shocks and local spillovers are accounted for. We alternatively consider the components of government expenditure both as share of GDP and as share of total government expenditure. This distinction matters because the effect of the single components as share of total government expenditure should be considered as the effect of a reallocation of resources towards a given spending component, since the size of government intervention does not adjust. In addition, also this chapter includes the GBC in the regressions in order to obtain estimates of the response of private investments to the various combinations of spending components and methods of financing.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.