Effect of Governance on Educational Performance in Nepal

Good governance often seems to have accelerated educational performance. Stepping onto the contribution of governance to the education, this paper examines the effect of Worldwide Governance Indicators produced by Kaufmann et al. (1999) on Educational Performance (EP) of Nepal during the years from 1996 to 2018. The six indicators of WGIs: political stability and absence of violence, government effectiveness, voice and accountability, regulatory quality, control of corruption and rule of law are used as independent variables, and the educational performance (student learning achievement and education index) as a dependent variable. The results, based on the data collected from the secondary sources, derived from multiple-line graphs and the regression model shows that the majority of WGIs insignificantly explained the educational performance across the years. One indicator namely government effectiveness is found as a negative significant predictor of EP. The findings of this study suggest to reform in the existing level of WGIs for the better educational performance.


Introduction
Effective governance is essential factor for the development of education as it is assumed as an entry point to educational development (Lewis & Pettersson, 2009). Further, they illuminate that the practice of governance can make education system functional to raise the educational performance in effective and efficient way. Governance effectiveness is assessed through different indicators that are largely similar to all organs of a state and worldwide. In this study, the Worldwide Governance Indicators (WGIs) model produced by Kaufmann et al. (1999) that consists of six dimensions of governance (voice and accountability, political stability and absence of violence, government effectiveness, regulatory quality, rule of law and control of corruption) has been used to assess the effect on educational performance (World Bank, 2020).
While considering the effect of governance on educational performance, it is revealed as a predictor of socio-economic development of a country such as health, education, economy, livelihood of people and so on. However, there is no uniform results, for example, WGIs are found positive impactful factors on development indicators such as education, Gross Domestic Product Per Capita (GDPPC), human development index (Ahmad & Saleem, 2014;Canfield, 2011;Han et al., 2014;Zaman, 2016) whereas some researchers like Vinayagathasan and Ramesh (2019), Zaki (2016), Briguglio (2016), and Abu-Ismail et al. (2016) showed the negative correlation between some indicators of WGIs such as voice and accountability, political stability, rule of law or control of corruption, and average year of schooling or GDPPC. The mixed results on the relation between governance and the development indicators indicate it as a debatable issue. However, the majority of researches agree with the positive impact of governance on the overall development of a country. Based on these evidences, this study theoretically assumes that good governance positively explains the educational performance and following this assumption, this study is led by the question: How has the effect of WGIs on EP in Nepali education context. rules and principles. Later, in 1980s, the governance practice has been shifted from classical approach to new public management that injected the principles of competition and private sector management in resource allocation and service delivery within public services including health and education. After a few decades, in about 2000s, new public governance took place that places the citizens as the co-producers and service delivery at the centre of governance framework. Now, the governance approach has been shifted to the new public service guided by democratic theory, and emphasizes on citizens, community and civil society (Robinson, 2015). In connection to governance, Ferlie et al. (2009) highlight the concerns of governance such as public administration with its implementation approach, decision making style, power diffusion, responsiveness and accountability in public service delivery. From this brief review on governance, the governance mainly revolves around public policy implementation, decision making style, control mechanism for allocating resources and public service delivery.
Considering the fundamentals of governance, the World Bank Group reports aggregate and individual governance indicators that includes six indicators. The first indicator, voice and accountability, captures the perceptions towards the government representativeness, a freedom of expression or association of a country's citizen whereas the second, political stability and absence of violence refers to the measures that a government is not affected by politically motivated violence. The third indicator, government effectiveness, concerns with government's capability of developing policy, implementing them effectively in public server delivery, and the fourth, regulatory quality, captures the sense of effective policies and practice collaborating with the private sector. The fifth indicator, rule of law, indicates the extent of citizens who are confident that their governance agents follow the norms, values and rules in day-to-day activities, and last indicator, control of corruption, is measured in terms of (mis)use of authority (Zaman, 2016;The World Bank, 2020), Regarding educational performance (EP), it includes several components. In education research studies, it is measured in terms of several elements such as access, enrolment, dropout, completion rate, student learning achievement, cost per student, average years of schooling, graduation, employability, and so on (Education Review Office Journal of Education and Research, Vol. 11, No. 1, 2021 Programme for International Student Assessment (PISA) surveyed the performance of education systems in terms of student learning achievement that is the learning outcome achieved by students (OECD, 2010), whereas ERO of Nepal includes the regularity, repetition and drop-out rate of students, learning achievement, consumer satisfactions, teachers' job satisfaction are considered as the educational outcomes or performance for assessing school performance evaluation (ERO, 2019). Similarly, United Nation Development Programs includes education index as one of the dimensions of Human Development Index (HDI) which is calculated based on the mean years of schooling and the expected years of schooling (UNDP, 2019). Enders et al. (2012) states, in the context of European countries, for system performance access, mature learners, graduation, employability, international student mobility, research output, capacity to attract funding and cost effectiveness are selected as performance of higher education. In this study, EP is specified by the education index assessed by UNDP (2019) and the student learning achievement in terms of the results of school leaving examination (SLC) and school education examination (SEE) to be conducted by the Office of the Controller of Examinations (OCE) and National Examination Board (NEB), Nepal.

Educational Governance Practice in Nepal
The governance is directly or indirectly related to the process of service delivery of a state or its organizations to the people. In the education of Nepal, some studies have critiqued on the practice of governance. For example, Chaudhary (2019) reviewed some legal documents such as constitution, local government operation act, and similar sectoral legislations, and portrayed the views of some local governors. His study indicates the inadequacy of laws and policies, lack of coordination between three layers (state, province and local) of government are the challenges for good governance. Although local government is given the utmost level of decentralization such as devolution of power from federal (central) level to local level, he found that the local authorities were still seeking orders and directions from central level if any issue occurs. This study indicates the lack of smooth functioning of local government and also created a dilemma for the researcher that whether the local authority has low level capability to take decision and implement them or they have not been given the adequate decisive powers. In connection to the responsibilities of local government such as merger of schools, transfer of teachers, management of teachers, and textbook distribution, Neupane et al. (2018) state that local level governance practitioners have faced the challenges on the governance practice due to the continued assertion of decentralization, slow devolution of power to local authority, and jurisdictional confusion between the teachers and local government officials.
Some researchers like Ghimire (2015) and Truex (2011) have revealed that even more than the past, the governance is found less effective, especially in education sector. Ghimire (2015) has critiqued the Nepal's education system as most corrupt sector presenting the evidence reported by the Commission for the Investigative Abuse of Authority (CIAA). According to the CIAA's report of 2013/14, largest numbers (13% out of all cases) of corruption-related cases were registered and reported education sector having maximum corruption and irregularities. For example, hundreds of fake schools 'Jhole Bidhyalaya' were registered, and in their names, millions of rupees (Nepali currency) had been embezzled on different expenditure-headings. Truex's (2011) survey study of the residents of Kathmandu districts identified that there was unacceptable level of practice of bribery but petty corruption, gift giving, and favouritism were found as the forms of corruption. Further, the study suggested to improve the access to education because more educated people were found less accepting attitude towards the range of corrupt behaviours.
Likely to control of corruption, accountability is also an important dimension of governance. In Nepali education system, review of student learning achievement, social audits, public hearing, parents' participation or networking are common indicators to assess the accountability. Evidence tells that in education sector, there is weak level of accountability. The NASA report 2018 shows the understanding performance in selected subjects (Mathematics and Nepali) of grade five, comparatively in public schools, and it is not significantly different from the NASA's results of 2012 and 2015 (ERO, 2019). Neupane et al. (2018) mentioned in the report that parents through school management committee and parents-teacher association have been working directly and indirectly to assess the school performance, control school management and maintain the school authority and communication with the school authority and teachers staff. Indicating under-performance of local governments, the study suggests the teachers to be professional in their occupation and detach it from politics.
According to the World Bank's report of 2018 on WGIs, the percentile rank of Nepal is below 50. The ranks of Nepal in voice and accountability, political stability and Journal of Education and Research, Vol. 11, No. 1, 2021 absence of violence, government effectiveness, regulatory quality, rule of law and control corruption are 39. 41, 13.81, 9.62, 23.08, 33.65 and 27.40 respectively (The World Bank, 2020). Out of the indicators, the rank in voice and accountability of Nepal is better as compared to others, whereas the rank of government effectiveness appears to be poorer. These all evidences indicate the unsatisfactory level of governance indicators in Nepal

Educational Performance: Nepal Context
In this study, educational performance includes two dimensions: education index and student learning achievement in aggregate. Educational performance is considered to examine the one-to-one relation with WGIs from the years of 1996 to 2018. The UNDP's (2019) human development reports include people's education as one of the indicators of HDI. Education index is examined in terms of mean years and expected years of schooling which is indexed to assess the HDI (UNDP, 2019). Mean years of schooling shows the average number of schooling spent by the adults of age 25 years and above, and the expected year of schooling is estimated as the "Number of years of schooling that a child of school entrance age can expect to receive if prevailing patterns of age-specific enrolment rates persist through the child life." (UNDP, 2019, 303). As reported in the human development reports, Nepal is ranked at 147th position of education index out of 189 countries in the year 2018 which shows the poor education index (= 0.498) between the highest education index (= 0.946) of Norway and the least index (= 0.246) of Niger worldwide. Student learning achievement, as another component of educational performance, is a student's ability (knowledge, skills and behaviour) gained after the completion of a grade or level. It is assessed using different tools and techniques. Standardized test, teacher-made test and other formal or informal assessments of evaluation are used to assess the student learning achievement for summative or formative evaluation or both purposes. As provision in national curriculum framework 2019, there is continuous assessment-CAS system from class one to seven, and from grade 8 to 12, in formative evaluation, homework, class assignments, project work, group works, unit or monthly tests, and extra-curricular activities are the major means for evaluation, whereas in summative evaluation, terminal and final examinations are administered that include both theoretical and practical examination (Curriculum Development Centre [CDC],

Effect of Governance on Educational Performance | 103
Journal of Education andResearch, Vol. 11, No. 1, 2021 2019;CDC, 2014). As stated in policy document, the final examination of grade 8 is conducted by local level; grade 10, province level; and grade 12, national level. An autonomous body, ERO, has also been established at central level that conducts national assessment of student achievement for the baseline and compare the achievement with the previous results. In this study, assessment for summative evaluation for school level education is used for the analysis.
As aforementioned, in Nepal, till the date, the School Education Examination (SEE) is administered at the grade 10 by external public agency, NEB. From the year 2015, the result of this examination is evaluated in letter grading system based on cumulative grade point average (CGPA) 4. In this system of evaluation, students are not categorized as 'pass' or 'fail'. However, as stated in the notice of Curriculum Development Centre on August 18, 2020, Grade Point Average (GPA) D+ in each compulsory subject with CGPA 1.6 is the eligibility criteria for studying in grade XI (CDC, 2020). In this study, CGPA 1.6 is taken as the minimum grade point for finding the number of pass students from the year of 2015 to 2018. Before this grading system, the examination, called School Leaving Certificate (SLC) examination, was used to be administered by the Office of The Controller of Examination (OEC) and there was the system of scoring out of 100, and the minimum pass percentage was 32% in each subject. The record of the OCE shows that the pass percentage from 1996 to 2014, was in the range of 31.62% to 68.47%, where 50% of the total students, in average, would fail every year. When CGPA 1.6 is assumed as the minimum criteria for pass result, the pass rate from the year 2015 reached above 80%. The result in grading system shows the drastic improvement in the pass rate as compared to the results before the year 2015. This incomparable results between scoring system and grading system raised the issue that whether there is greater improvement in school system or there is an error in evaluation criteria. This might be due to technical reason like aggregate grading without considering individual subject grading or improvement in education quality that may be the subject of future research.

Research Methods
In this study, the effect of Worldwide Governance Indicators (WGIs) on the educational performance (EP) of Nepal is examined. The set of WGIs is defined as independent variable that includes six dimensions: voice and accountability, political Journal of Education and Research, Vol. 11, No. 1, 2021 stability and absence of violence, government effectiveness, regulatory quality, rule of law, and control of corruption. The data for the governance indicators were collected from the Databank of The World Bank Group (2020). Regarding the source of WGIs, it is stated in the policy paper that WGI project combines the several hundred individual underlying variables into six indicators (Kaufmann et al., 2010). The data are obtained from 31 different sources such as survey respondents, government organizations, commercial business information providers, and public sector organization worldwide, and the data sources are rescaled to create the six indicators of WGIs. The scale of the data is rescaled into standard normal units that ranges from -2.5 to 2.5. The value of WGIs greater than 0.00 is said to have positive indicator, and if it is less than 0.00, then it is said to be negative indicator (WBG, 2020). For the years: 1997, 1999 and 2001 the database for WGIs were not available, for these missing data, as stated by Pratama et al. (2016), mean imputation as conventional method is used to fill the missing data. In descriptive statistics, WGIs and EP were analyzed using multiple-line graphs as it is easier to observe the trend of the variables for the years. For inferential analysis, the data on WGIs and educational performance both are in different matric scales. For examining the effect of WGIs on EP, as the ten observations for each independent variable in the sample observations (Miller & Kunce, 1973;Halinski & Feldt, 1970, as cited in Bartlett et al., 2001) is satisfied, linear regression model is used. For the testing the assumptions for the regression, as it is based on the time series, Durbin-Watson test for autocorrelation, Variance Inflation Factor (VIF) for multicollinearity, plots (P-P plot and Scatter plot) for normality and homoscedasticity are used. Before the statistical analysis, the data were standardized for rescaling into the same metric data as the standardized data are identical to the original distribution of the real data. As stated in

Effect of Governance on Educational Performance | 105
Journal of Education andResearch, Vol. 11, No. 1, 2021 Singh (2007), the effects of the independent variables are compared in standard deviations, instead of the real units, within the model.

Results
In the study of trend analysis of WGIs and EP, the following multiple-line graphs are drawn using SPSS. The figures 1 and 2 show the multiple-line graphs for WGIs and EP from the year of 1996 to 2018.

Figure 1
Lines Graph for Worldwide Governance Indicators across the Years from 1996-2018 Figure 1 indicates that the trend of all WGIs lines graph fall below '0.00' which shows the negative values of WGIs. Out of six indicators, political stability and absence of violence seems inconsistent and it is in the vulnerable situation up to 2004/05 and it has been increasing since then. The situation of voice and accountability is comparatively better than others since 2005/06. One more surprising trend of the governance indicators is found in government effectiveness that has been decreasing from the beginning to the year of 2018. This similar situation is followed by the practice of control of corruption, rule of law and regulatory quality. In overall observation of the WGIs line graphs, between the years 2003/04 to 2013/014, the trend of all WGIs seems weak in the period of transition from constitutional monarchy multiparty democracy to federal republic democracy. After the establishment of federal republic democracy, there was the situation of instable governments, dilemma in interim-constitution in power allocation, power struggle to create space for diverse groups and the constitutional assembly elections. After the year 2013/2014, all lines graph tends to be similar range and increase towards the positive direction that indicates the symptoms of revival of WGIs better than the past.

Figure 2
Lines Graph for Educational Performance across the Years from 1996-2018 Figure 2 shows the line graph for educational performance. The line graph has been increasing from the year 1996 to 2018. In the period of 1996 to 2005, the performance is very low. After the end of armed conflict and advent of federal government, the result is comparatively higher for next three years up to 2008 and goes downwards up to 2012 and again turns upward. The line graph has inclined upwards at higher degree from the year 2015 that might be the criterion changed from scoring to grade point.
The following section examines the effect of WGIs on educational performance in Nepal. First, the assumptions aforementioned are examined that whether they were satisfied or not. Then, the regression analysis is made if the assumptions were satisfied.

Normality and Homoscedasticity
For the educational performance as dependent variable, P-P plot and Scatter plot are used for testing their normality and homoscedasticity. Normality P-P plot of regression standardized and the scatter plot for homoscedasticity are presented in the figures 3 and 4:

Figure 3
Normal P-P Plot of Regression Standardized Residual for Educational Performance

Figure 4
Homoscedasticity Scatter Plot for Educational Performance The residuals (different between standardized value of education performance) tend to be normally distributed as these residual points cluster to diagonal normality line plotted in the figure 3. In figure 4, the residual points tend to scatter about the horizontal line that shows not very violated level of homoscedasticity.
Journal of Education and Research, Vol. 11, No. 1, 2021  The value of R (=.839) (multiple correlation coefficient) shows very high level of correlation between dependent and independent variables. The adjusted R square (=.593) indicates 59.3% of the variation in educational performance (dependent variable) is explained by the six indicators of WGIs. As stated by Singh (2007), the value of adjusted R square more than 50-75 percent shows that the regression model is at 'good' level to use for analyzing the results.
The data is based on time series, autocorrelation for assessing the degree of similarity between the given time series of 23 years is tested by Durbin-Watson. The value of Durbin-Watson is .709 that does not seem much more violated the assumption of 'No autocorrelation' although it lay outside the region of 'no autocorrelation' of 1.5 to 2.5. Further, whether model fits the analysis or not is examined by the following ANOVA table: Journal of Education and Research, Vol. 11, No. 1, 2021 The result of ANOVA presented in Table 2 shows that the independent variables statistically and significantly explain the dependent variables as the p-value (=.001) is less than 5% of level of significance. The regression model significantly fits for the regression equation (Singh, 2007). As goodness of fit for the equation and significant association of the WGIs and EP, the effect of WGIs on EP (major concern of this study) is examined as follows: In Table 3, the VIF values are not above 10, and as stated by Landau and Everitt (2004), multicollinearity among the independent variables is not the case of concern. Out of six indicators of worldwide governance, all the indicators except government effectiveness are found insignificant predictors of educational performance as the Journal of Education andResearch, Vol. 11, No. 1, 2021 values of p = .414, .514, .569, .250 &.1159 are less than the 5% = .05 level of significance whereas the result shows that the government effectiveness negatively explains educational performance (p = .030 < .05). The standardized coefficient (b) (= -.656) shows that when there is unit change in standard deviation of government effectiveness, educational performance changes by -.656 units of standard deviations. It means that even at the decreasing state of government effectiveness, educational performance can be increased.
In a nutshell, out of six indicators of WGIs, five indicators insignificantly explained the educational performance across the years from 1996 to 2018 in Nepal whereas one of them (government effectiveness) is found significant negative predictor of educational performance.

Discussion
The mixed results regarding the effect of WGIs on educational performance are comparable and contrastable to the similar past studies. The majority of the findings of the past researches contradicted to the findings of this study. For instances, Zaman (2016), Ahmad and Saleem (2014), Han et al. (2014) and Canfield (2011) have revealed the positive, predictive and significant relation between worldwide governance indicators and different dependent variables such as economic growth, human development, domestic product, educational outcomes or country growth. However, some researchers found some dimensions of WGIs that negatively correlate to educational outcomes and others such as government effectiveness with GDP in South Asian countries, and with higher education enrollment in Latin American and Caribbean counties in the years from 1996-2012 (Zaman, 2016). Vinayagathasan and Ramesh's (2019) study also showed the negative relation between rule of law and GDPPC in the case of Sri Lanka from 1996 to 2016.
In practical sense, bad governance is not the requirement for the progress in educational performance (Briguglio, 2016) that contradicts the general notion. The surprising finding of this study is also that a single indicator of WGIs (government effectiveness) is a negative predictor to educational performance of Nepal. For this unexpected result, two major conditions are discussed in the literature. The first is diminishing marginal effect of governance (DMEG) and the second might be the effect of unobserved variables. By the law of DMEG, a little effort to improve governance would have higher effect on growth rate in low-income countries if they are at the starting phase of development. The second condition, as stated by Abu-Ismail et al. (2016), is the effect of other variables like interventions of international organization other than governance. They found the significant progress in the health indicator of HDI in the developing countries with a low level of governance in majority of Arab countries.
In the context of Nepal, above two reasons might cause the result of negative relation between government effectiveness and educational performance. As compared the educational performance in terms of student learning achievement (SLC/SEE results) and education index, in the year 1996, the pass percentage of SLC/SEE result was 36.52% and EI, 0.31, and in 2018, these indicators raised to 87.64% and .50 respectively that indicate developing phase of growth. Secondly, in Nepal, many national or international organizations besides the governmental organizations such as UNESCO, Save the Children, Plan Nepal, and many others are actively working to improve educational performances. However, these reasons are not the strong support for this finding and require further research.

Conclusion
This study examines whether there was the effect of governance on educational performance in Nepali context. The trend of governance practice in terms of GWIs has been gradually improving in the late years during the study period of 1996 to 2018 however the state of governance is unexpectedly low. The effect of government effectiveness is even worse which raised the questions on the effectiveness government's capability of developing and implementing policy and providing the public services. Despite the low level of governance practice, the country has made substantial growth on educational performance. Student learning achievement and education index both have been improved during this period. The results of this study on these two differently directed variables (WGIs and EP), the author concludes that the governance measured in terms of WGIs has little on educational performance of Nepal during last 23 years. Further, the findings of this study suggest to reform in the existing level of all worldwide governance indicators for the better educational performance as these indicators are either insignificant or negatively significant.