The subject of this dissertation is the quantification of the different types of economic policy uncertainty, the interrelationship among them, and the analysis of their macroeconomic effects. Knight (1921) defined the concept of economic uncertainty as the impossibility of quantifying the probability of certain future events. Economic uncertainty is a latent variable. It has attracted special attention from economists during the economic crisis from 2008, as uncertainty could be the missing link in macroeconomic modeling and prediction of economic downturns. The COVID -19 pandemic has also increased researchers' interest in quantifying economic uncertainty. Quantifying a latent variable is not an easy task. There are three different methods for measuring economic uncertainty in the literature. First, Jurado et al. (2015) compute a measure of macroeconomic uncertainty. Their indicator of macroeconomic uncertainty captures the aggregate variation in uncertainty across many macroeconomic time series. The second approach to measuring economic uncertainty is forecast disagreement among economic agents. In uncertain times, forecasts by professional forecasters will differ widely, while in certain periods forecasts will be approximately the same. The final method of quantifying uncertainty is based on media reports. Baker, Bloom, and Davis (2016) quantified the economic policy uncertainty index using media articles from the most widely read newspapers in the United States. The construction of the index is based on the frequency of articles that contain at least one word from all three categories: economy, policy and uncertainty. For each of the three categories, there are predefined keywords for searching the newspaper articles. The authors apply a vector autoregression model and show that aggregate economic policy uncertainty innovations anticipate a decline in output, employment, and investment. They also construct an aggregate indicator of economic policy uncertainty for a group of eleven countries. In addition to the aggregate indicator, Baker, Bloom, and Davis (2016) quantify eleven different types of economic policy uncertainty related to taxes, government spending, fiscal policy, monetary policy, healthcare, national security, regulation, financial regulation, sovereign debt and currency crises, entitlement programs, and trade policy. On the other hand, Sorić and Lolić (2017) quantify the aggregate economic policy uncertainty indicator for Croatia using the frequency counts from media articles that contain at least one keyword from all three categories: economy, policy, and uncertainty. This dissertation is a continuation of the research shown previously. In this dissertation, the types of economic policy uncertainty for the Republic of Croatia are defined and the types of economic policy uncertainty for the United States are methodologically improved. It is important to quantify the types of economic policy uncertainty, as uncertainty can affect individual sectors and individual economic agents to varying degrees. Uncertainty can significantly affect decision making about personal consumption, investment, and the implementation and efficiency of economic policy. Using newspaper articles from the six most widely read newspapers in Croatia and applying the methodological approach of Baker, Bloom, and Davis (2016), eleven types of economic policy uncertainty were constructed for Croatia. Namely, economic policy uncertainty related to taxes, government spending, fiscal policy, monetary policy, healthcare, national security, regulation, financial regulation, sovereign debt and currency crises, entitlement programs, and trade policy. A similar approach is used for the U.S., using an expanded set of keywords. An uncertainty indicator should map the main events related to any economic policy of interest and behave countercyclically. Any shock in the constructed types of economic policy is therefore explained by important economic events it depicts. Also, the correlation with industrial production is negative for all types of economic policy uncertainty in Croatia and majority for the U.S. For the eleven quantified types of economic policy uncertainty in both countries studied, the interdependence of all types is estimated using the spillover index defined by Diebold and Yilmaz (2009, 2012). The analysis showed that the largest net transmitter of uncertainty shocks in both countries is fiscal policy. Moreover, the dynamic spillover analysis in Croatia shows that the total uncertainty spillovers vary between 65% and 82% in the analyzed period. Further, all shocks in the dynamic total spillover of economic policy uncertainty can be attributed to various economic and political shocks, in particular the 2008 economic crisis, the parliamentary elections, and the recent coronavirus crisis. For the U.S., the dynamic analysis has also shown that the estimated parameters are not fixed during the period analyzed. The random forest model was applied to investigate the intensity of the impact of the eleven types of uncertainty on a set of macroeconomic indicators. The chosen macroeconomic variables are industrial production, employment, investment and GDP. The analysis is performed separately for Croatia and for the U.S. Two models are estimated for the U.S. The first one includes the economic policy uncertainty categories of Baker, Bloom, and Davis (2016), while the second model uses the modified economic policy uncertainty indicators constructed in this dissertation. A robustness check is also performed, replacing the dependent variables with residuals from the additional estimated macroeconomic models. The results of the random forest model for Croatia are as follows. The most important variables explaining variations in industrial production and GDP are government spending, entitlement, and trade policy uncertainty. Second, trade policy uncertainty contributes the most to the explanation of employment, while the robustness check concludes that the most important variable is debt and currency crisis uncertainty. Finally, trade policy uncertainty explains most strongly the fluctuations in investment, while fiscal policy uncertainty is the most important in explaining the residuals. The empirical results for the U.S. have shown that the most important variable in explaining industrial production, GDP, investment, and related residuals is trade policy uncertainty. In addition, the most important variable to explain employment is healthcare policy uncertainty. The above conclusions indicate that economic policymakers should act countercyclically and use appropriate and timely instruments to mitigate the potential negative consequences of the highlighted types of economic policy uncertainty on the economy.