Institutions as the Main Determinant in Economic Growth

The studies on human capital and technological progress have given incredible insights on how countries in the world differ from one another. Yet there are more than those two reasons to account for differences among countries. There is a third reason why a country would differ in terms of its economic development progress, namely institutional factors. Hence developing institutional indices would give a deeper explanation than a mere theory. On the other hand, we can corroborate the institutional index with the general theory that low-quality institutions will impact an economy negatively. This study seeks to broaden the understanding of causes of economic growth by incorporating institutional index into a semiendogenous growth model and finds a relationship between that index with human capital and technological progress.


Introduction
A question of interest for most of the macroeconomists is a question related to the cause of the difference in economic growth among countries in the world.According to Acemoglu & Robinson (2008), economic growth is related to people's ability to accumulate human capital, physical capital, and technology.Acemoglu & Robinson (2008) further summarize the causes of the difference in economic growth to only 2, namely the proximate cause and fundamental cause.Proximate cause or the most active and dominant cause is human capital, physical capital, and technology.While the fundamental cause is the most basic, and it will became the major cause of the resulting proximate cause.So if we want to get a satisfactory answer from the question at the beginning of the paragraph, then we should focus on the fundamental cause.
The fundamental cause is the most basic factor and the cause of differences in every country in the world.Institutions are fundamental causes that cause differences in world economic growth (Acemoglu & Robinson, 2008).Institutions shape how society behave and react to certain challenges in their lives.Institutions command a society to react by creating certain rules and regulations as guidelines.These rules and regulations are sometimes in the written formal code but most of the times they need not be written.Institutions are reflections of society and the people who live in therefore different countries possess different institutions.Therefore institutions are fundamental to every country as they dictate paths to where a country might progress.Whether it is on the right path or on the wrong one, clearly rely on institutions at play.Different countries possess different qualities of institutions.Economic institutions determine incentives and constraints for economic actors and contribute to shaping the output of the economy.Economic institutions involve social choices in which social choice will vary between individuals so that these social choices will lead to conflict.Those who have a greater political advantage will ultimately win the conflict.Developed countries with high economic growth are supported by innovation and growth-oriented institutions.So the difference in economic growth in developed countries within developing countries lies in the quality of the institution.Therefore optimum human capital and technological advances, as drivers of economic growth, must go hand in hand with good institutions.This is interesting because if we understand and know the quality and position of these institutions from institutions in developed countries so we can map the problems and catch up quickly with a higher rate of economic growth.By knowing our position in the global map of institutions, we can map and list our strength and weaknesses and later improve them.The rapid growth of the stock of knowledge and technological progress begin to flourish once good institutions are established.Advanced economies all have good institutions in place.Good institutions can foster the growth of the stock of knowledge and technological progress by accommodating all possible knowledge spillover from advanced and other countries.The difference in the rate of economic growth between countries is not merely the issue of capital accumulation, the role of technology, and human capital; there is a role of institutions in it (Acemoglu et al., 2001).They suggest that in certain cases, institutions are the main determinants of economic growth.
With recent development in semi-endogenous growth model and the availability of data on institutions, it is imperative to develop an extension to semi-endogenous growth model with attention towards institutions.Semi-endogenous growth model emphasizes ideas creation where human capital and technological progress are the key ingredients.There must exist a relationship between institutions and ideas creation aspects.The first step in extending a semi-endogenous growth model can be directed toward building an institutional index.This study attempts to employ PCA methodology popularized by Filmer & Pritchett (2001) in building an institutional index.Later, a relationship between institutions and ideas creation is established.The previous study that has attempted to elaborate institutions in their analysis is of Hall & Jones (1999).The big difference with our study lies in their methodology in constructing the index.Hall & Jones (1999) use simple mean techniques of several institutions indicators developed by Knack & Keefer (1995), while our methodology is based on PCA.Since institutions influence knowledge directly and thus influence the rate of economic growth, we then build an institutional index in relation to the growth of knowledge.The technique used to calculate the growth of knowledge follows Jones (2002).The next section will briefly explain institutions and PCA methodology before we come to calculate the institutional index.

Methods
This study will use some indicators of the International Country Risk Guide (ICRG) database between 1983 and 2013.Indicators used in this study are perceived to have direct impacts on the stock of knowledge and technological progress.The choice of the time span is merely a case of data availability.The ICRG data that is utilized in this study consists of: Investment Profile measures risk factors in the business.Risk factors in the business include contract cancellation factors, the rate at which investors can recover their capital (repatriation), and the level of government delay in making payments to investors.Internal Conflict measures the factors of political violence and its influence on government.Corruption assesses the extent of corruption in the order of the political system.Law and Order measure the level of strength, independence, and fairness of the legal system.On the point of view of the government and institutions in relations with knowledge and technological affairs then the greater the risk signals the more likely knowledge and technological affairs receive less attention or become the priority.On the point of view of an investor then the greater the risk the less likely an investor would invest in the projects.
This study includes 100 countries, after that the countries can be grouped according to values of S and growth of A t .S, and A t are institutional index and growth of knowledge consecutively.We then plot this value on quadrant graphs that will give us 4 quadrants of countries.The use of this classification method is due to the number of countries involved but also to give better views on the current position of each country.This study is also new in terms of using 100 countries in the analysis.The original Jones (2002) involves only 6 countries.We will combine the concept of growth of A t as is explained in Jones (2002) to find correlation among sources of growth.The number of studies involving many countries is expected to provide a more general and comprehensive picture and can represent the entire sample well.
To avoid collinearity bias, this study used Principal Component Analysis (PCA) to form a composite index.PCA is used to describe the variance-covariance matrix structure of a set of variables through a linear combination of these variables.In general, the main components can be useful for the reduction and interpretation of variables.Let's say there are p variables consisting of n objects.Suppose also that from p variable, k main component is made (with k ≤ p) which is a linear combination of p of that variable.K, the main component, can replace the p-variables that makeup without losing much information about the whole variable.Generally, PCA is an intermediate analysis that means the main component results can be used for further analysis.
To calculate the impact of social infrastructure or institutions we will use a composite measure defined as "the sum of the weighted components of the political risk measure of the International Country Risk Guide".The index is based on the rating of the ICRG on the 6 components as below: (i) 12 points for each variable that includes investment profile and internal conflict; (ii) 6 points for each variable including corruption and law and order; (iii) 4 points for each variable that includes bureaucratic quality.Socioeconomic conditions variables are not included, because these variables are related to economic performance.So, it has a great possibility to influence perceptions of the institution as described by Jellema & Roland (2011).His replacement uses an additional open trade.So the PCA model used is: (12) Where: Some of the criteria used to find the best dynamic model or GMM model are: First, Not biased.Estimators of pooled least squares are biased upward and estimators of fixed effects are biased downwards.An unbiased estimate is in between.Second, the instrument must be valid.The valid meaning is if there is no correlation between the instrument and the component error.This validity is checked using the Sargan test.The null hypothesis of the Sargan test states that the instrument has no problem with validity (valid instrument).The instrument will be valid if the Sargan test cannot reject the null hypothesis.If the result of the AB-GMM method indicates the instrument used is invalid, then the SYS-GMM method is used.Third, the estimation result must be consistent.An autocorrelation test on the GMM approach is used to determine the consistency of the estimation results.The consistency properties of the estimators obtained can be checked from Arellano-Bond statistics m 1 and m 2 , which can be calculated automatically on some software.The estimate will be consistent if m 1 denotes the null hypothesis is rejected and m 2 indicates the null hypothesis is not rejected (Arellano & Bond, 1991) Result and Discussion Islam (1995) mentions the importance of the institutional role in explaining differences in economic growth.Research on the determinants of differences in economic growth and income between countries can be grouped into 3 broad theoretical groups.The first group of theories is a group that focuses on factors of input to production processes, such as physical and human capital, and technological advances that support economic performance.Solow (1956), Lucas Jr (1988), Romer (1986Romer ( , 1990)), Grossman & Helpman (1991), Jones (1995aJones ( , 1995bJones ( , 2002)), Segerstrom (1998) had started the discussion about this topic.Endogenous growth models fall into the first group.The second theoretical group is the focus group on location and geographical location where certain characteristics will support the economy to reach the highest level of growth while the location or other location is less supportive.Sachs (2001), Gallup et al., (1999), and others had conducted the study from this group.A third theoretical group is a group that focuses on institutions as a driver of economic growth.North (1991) had pioneered the study in the third group.
Good institutions foster the growth of knowledge.Jones (2002) argues that the engine of economic growth is the creation of ideas.Jones (2002) seeks to explain the stagnant rate of growth in the United States during 1950 -1993 and concludes that much of the growth is attributable to the growth in ideas (almost 70%).Jones (2002) mentions that the differences among economies are endowment and allocation.This creates opportunities in extending Jones (2002) by incorporating institutions into the model.Good or bad institutions can be considered as endowment while effective or ineffective allocations are results of good or bad institutions.Jones (2002) proposes to calculate the accumulated knowledge with the following equation: (2) In Nugroho (2018), equation 3 has been modified to become: (3) The channel we will use to incorporate institutional index into Jones modified model is through variable, S, which we limit its value to a maximum of 1.The reason for the maximum value is to that after a certain country reaches S =1 then it becomes what Jones (2002) explained in his research, an advanced country.As S ≈ 1, the equation (10) will return to its original Jones (2002) version, in equation ( 9).Jones (2002) assumes that countries in his model are all the same in level of ability to conduct research and to foster new knowledge.In our model, it then translates to S = 1.We allow for countries' variabilities in their ability to conduct research and to foster new knowledge.This is a novelty this study will offer.
Determination of the number of components in the PCA is done by searching for variables or components that are not correlated, independent of each other, but fewer than the initial variables.Although it produces a fewer number of variables, it absorbs most of the information contained in many more initial variables and can contribute to the variance of all variables.In PCA, the determination of the component refers to the eigenvalue value, indicating the amount of contribution of the component to the variance or diversity of all initial variables.In this case, if the eigenvalue value obtained is greater than one, then the component formed can be maintained.Otherwise, if the eigenvalue value is less than one, then the component cannot be used.
Table 1 presents the results of eigenvalue calculations for the formation of the corporate vulnerability index, the percentage of total diversity (Proportion) and the cumulative total diversity (Cumulative Proportion) capable of being explained by the diversity of the components formed.Based on Table 1, of the 12 components formed there are three components that have eigenvalue greater than one.Component 1 has an Eigenvalue of 2.784994, Component 2 of 1.255002, and Component 3 of 1.010487.Meanwhile, in Table 1 there is also a column 'Proportion' which shows the percentage of variance or diversity that can be explained by each component and column "Cumulative Proportion" which describes the cumulative of each component simultaneously.The magnitude of diversity capable of being explained by Component 1 is 46.42 percent.The diversity explained by Components 1 and 2 is 67.33 percent.The diversity explained by Components 1, 2, and 3 is 84.17 percent.Based on the eigenvalue of the three components greater than 1, and the cumulative percentage of the three components of 84.17 percent, it can be concluded that the three components can represent the diversity of the initial variables.Table 2 presents the component matrix that shows the magnitude of the correlation of each variable in the formed component, or loading factor.Based on Table 2 below, it appears that there are three factors or components that are formed from the six indicators of vulnerability.This shows that the three components are the most optimal amount to reduce the six original variables.
We can determine the Factor Equation by comparing the correlation value on each line within each component formed (see Table 2).We use the general form below to generate a factor equation.
Where expresses mean and is a standard deviation of indicators (component), .α expresses weight or loading factor of each indicator, in the first main component.If we combine the information from Table 2, with the above factor equation, we can get: Where: X 1 is bureaucratic quality; X 2 is an investment profile; X 3 is internal conflict; X 4 is corruption; X 5 is law and order; X 6 is trade-openness.
From the results, the most ideal in the determination of the institutional index is the first group.This is because the nature of the equation is non-negativity that means each indicator gives a positive contribution to the resulting index.After obtaining the index value of each country we then do rescaling of value between 0 -1.We use value 0 -1 to analyze institutional index where 0 is the minimum value of the institutional index while 1 is the maximum value.Before we come to the growth of A t , we calculate A t using the following equation: Thus the growth of A t is the log of two consecutive years or we can take a log of the above equation then differentiate with respect to time.We need to calculate the growth of A t in order to compare each individual country's institutional indices with the growth of A t .In doing so, we can easily make a conclusion about the relationship between institutional indices with the growth of A t .We can say that to measure productivity, we calculate the growth of A t and the the indices, the higher the productivity of a country will be.Hence, countries with higher productivity also translate to higher economic growth.Table 3 shows that OECD countries dominate index values with the first top 20 countries.It proves that OECD countries have better institutions compared to other groups of countries.Another interesting finding is Botswana that places no 27, just above Italy.Perhaps this is so because of the successful and continuous efforts of the Botswana government in eradicating corruptions.North (1991) defines institutions, as rules of the game in a society or in more formal definitions do humans that ultimately shape human relationships within the society create boundaries.In the case of Botswana, its government has done a great job in defining rules of the game in society hence translating to a better quality of institutions and economic growth.North (1991) argues that institutions are a major cause of economic development and have hypothesized that institutions play a role in both short and long term growth.As we can see most African countries lack good quality institutions resulting in lower economic growth.The rest of the African countries is at the bottom of the index.As was expected, most OECD countries lie in the first quadrant that is characterized by the value of S near 1 and the above average value of growth of At (See Figure 1).From the first quadrant, we can also find that Luxembourg (69) has the highest institutional index (0,9371) of all 100 countries (refer to Table 4).Another interesting finding is that Botswana, the only African country, made it to the first quadrant with the value of index 0,6393 (refer to Table 5).This can be related to the successful effort of the Government of Botswana in fighting against corruption in recent years.If this value is compared to that of Indonesia, then Indonesia still falls behind Botswana (0,6393 > 0,3797).Yildirim & Golkap (2016) says that institutional factors can increase or decrease productivity.To achieve high economic growth, the state must have institutions that encourage every organization within the country to engage in productive activities.In developing countries, the existing institutions prioritize distribution activities rather than production activities so that conditions leading to monopolies are created that ultimately inhibit productive opportunities.In addition to increased productivity, good institutions will increase efficiency and trust.Complete values of the institutional index of all 100 countries can be referred to Table 3.
The effect of institutional indexes is estimated using the GMM method, where the data used is non-OECD data because OECD countries are considered as the maximum limit of the multifactor productivity value.The estimation result shows that the model used is valid and consistent.This result is valid and consistent (See Table 6).Variables that have an impact on the growth of A t are human capital, multifactor productivity of the country, and multifactor productivity of advanced countries, institutional index, and growth of A t in the preceding year.Value of coefficient of the institutional index of 0.0310 describes that whenever there is a rise of 1% in the institutional index then it will raise the growth of A t by 0.03%.Value of growth of A t at lag 1 is less than 1 but slightly more than 0 that shows there is convergence among OECD countries being analyzed.This result is consistent with Siddiqui & Ahmed (2013) that suggest favorable institutions positively affect economic growth.There is a causal link between a cluster of good institutions and rapid ling run economic growth (Lin & Chen, 2011).Institutional is a key role in the process of economic development (Osman et al., 2011;Roy et al., 2014;Ahmad & Hall, 2017).According to the result, it is imperative that the government should pay more attention to institutional indices.The improvement of institutional quality can attract more foreign direct investment (Kandil, 2009).The success of institutions is largely determined by the degree of accountability and corruption (Sumanjeet, 2015).The institutional reforms to upgrade the quality of both political and economic institutions are crucial for the countries (Slesman et al., 2015;Rachdi et al., 2018).

Conclusion
PCA result shows that there is a difference in index values between each quadrant.The result shows that developed countries have a tendency of higher index value relative to other countries.Quadrant I characterizes developed coun t ries in which institutional index is higher than any other quadrant.The growth of A t is fairly high in this quadrant but not the highest.Institutional influence on economic growth is evidenced by the results of GMM where the influence of institutions has a positive and significant impact on economic growth.It proves that countries with higher institutional indexes have higher economic growth than those with lower institutional indices.GMM results also prove that human capital and multifactor productivity have a significant effect on economic growth that means economic growth is not only influenced by capital and labor but also influenced by human capital and multifactor productivity variable.
The government must redefine its definition of good institutions as being innovation and growth-oriented institutions.From the perspecti ve of innovations, we mean that institutions must provide ways and environment to cu ltivate new ideas.Besides that, from growth-oriented, we mean that institutions must actively seek new ways to improve available knowledge and technology.The government can start from very technical research and development institutions and later make ways to other institutions in the nations.The improvement of the institutions that can foster and advance knowledge and technology are prioritized.Improvement can take many forms from increasing budget, increasing human capital involved, or creating conditions that can sustain continuous research and development.The conditions that sustain research and development can also be further supported by means of law and regulations.

X 5 :
Law and OrderX 6 : Trade Openness Panel data analysis can be used in dynamic models in relation to dynamic dynamics of adjustment.This dynamic relationship is characterized by the presence of lag of the dependent variable among the regressor variables.