Factors influencing life expectancy at birth in japan from 1970 2017

INTRODUCTION Health is one of the most priceless assets of human being has It permits us to fully develop our capacities Whenever this asset erodes, it indicates physical and emotional weakening, causing obstacles in the lives of people Therefore, human beings have continuously sought to improve their skills and to reach a life that is more and more dignified; for this reason, improvement in health has always been, today as in the past, one of the most important social objectives (AESS, 2015) To assess this regard, it can’t help mentioning the life expectancy at birth index It’s the determinant of HDI, the human development index, a synthetic measure reflecting human development in terms of income, knowledge, and health and this index assesses the level of socio-economic development of countries and territories Therefore, it is crucial to identify the factors which contribute to the health of the population in general and the life expectancy in particular The information on the nation’s health status helps policymakers and practitioners in their search for cost-effective mechanisms, providing health services and reallocation of health resources to optimize the gains from health expenditures (AESS, 2015) In almost all parts of the world, improvements in health and sanitation conditions, better living standards, higher educational attainments, and advancements in medical technologies are enabling people to live longer Compared to other countries, Japan leads the world (V Yong & Y Saito, 2009) , of which the position remained the highest life expectancy at birth index country, and even regularly increasing year by year Historically, there have been a lot of scientists, researchers, experts implementing these relevant to this topic On the other hand, Japanese people always make the whole world admired not only by the innovative inventions with extremely high applicability but also by cultural beauty, good manners, daily habits, and lifestyles All of these have created a living environment, the prerequisite for the highest life expectancy at birth index in the world However, there are not many studies that have used quantitative research methods to learn about the factors affecting life expectancy at birth in Japan to identify the level of influence of those factors in detail This study is one of very few studies which have investigated the effects of Nurses and midwives, School enrollment, tertiary, Population density, Communications, computer, etc - the development of technical science, Forest area, Daily smokers in Japan on life expectancy The data in previous studies are often taken from old data sources, not updated to the present time, 2019 That makes the results of research and analysis no longer accurate, causes bad influence to recommendations, suggestions after the investigation in the study Moreover, using time-series data method to clarify the relation between those indicators and is life expectancy even rarer Because of such shortcomings, it is difficult for other countries with the same level of economic development to draw lessons from Japan Other developed countries like China, America, Germany all have a much lower life expectancy index The Japanese can teach the world how to get a healthy and happy life from their wonderful culture Therefore, we chose this highly urgent topic “Factors influencing life expectancy at birth in Japan from 1960 to 2017” The objective of our assignment is to investigate the indicators and make recommendations, suggestions for other countries to improve the health living standard and develop the country LITERATURE REVIEW There are many studies that investigated the determinants of life expectancy and their dimensional influence on the issue Grossman (1972) investigated that inflation has a negative relationship with life expectancy, and household welfare is largely disturbed with rising prices Rogers (1979) first time gave a conceptual framework for life expectancy and income Robert Barro (1996) studied a panel of 100 countries from 1960 to 1990 and found that the growth rate of real per capita GDP was associated with longer life expectancy World Bank (1997) pointed out that there is a strong positive relationship between life expectancy and per capita income in the case of developing countries Davies and Kuhn (1992) found health intake and availability of food determine the health outcomes They concluded that investment in the health sector, social security programs would decide life expectancy Mahfuz (2008) focused on primary health care program as an important determinant of life expectancy On the base of his study, he concluded that there is a positive relationship between primary health care spending and health status Hill and King (1995) and Gulis (2000) investigated that education especially female education plays an important role in improving the overall life expectancy Williamson and Boehmer (1997) studied that educational status improves female life expectancy dramatically, their study is based on 97 cross-sections Anand and Ravallion (1993) investigated that there is a positive and significant relationship between life expectancy and per capita GNP, but it works through national income and public expenditures on health They mentioned that when public expenditures on health and poverty used as independent variables with per capita GNP the results are inverse to the first model Wilkinson (1996) explained after achieving a threshold level of per capita income, the relationship between life expectancy and standard of living to disappear and further increase in income is not attached to life expectancy gains He mentioned a direct relationship between health and income of the people at threshold level and there is no consistent relationship between them In a statistical analysis of life expectancy across countries using multiple regression, Tony Smith (2000) regressed Life Expectancy at Birth with a wide range of economic and social variables as below: ❖ Economy: GNP per capita ($US) 1995; GNP per capita annual growth rate (%) 1980-1995; Real GDP per Capita ($ Purchasing Power Parity) 1995; Average Annual Rate of Inflation (%) 1995; Country Development (1=developed, 0=underdeveloped) ❖ Population Characteristics: Urban population (% of total) 1995; Urban population annual growth rate (%) 1970-1995; Annual population growth rate (%) 1970-1995 ❖ Health: Public Expenditure on Health (% of GDP) 1990; 10 Physicians (per 100,000) 1993; 11 Contraceptive Prevalence (%) 1990-1995; 12 Fertility Rate (births per woman) 1995 ❖ Disease: 13 AIDS (per 100,000) 1996; 14 Tuberculosis (per 100,000) 1995 ❖ Access to Information/Technology: 15 Radios (per 1000) 1995; 16 Televisions (per 1000) 1995; 17 Newspapers (per 1000) 1995; 18 Telephone Lines (per 1000) 1995; 19 Electricity Consumption per Capita (kwh) 1995; 20 Commercial Energy Use per Capita (kg) 1994 ❖ Education: 21 Adult Literacy Rate (%) 1996; 22 School Enrollment Rate (%) 1995 - Combined first-second and third-level ❖ Environment: 23 Access to Safe Water (% of population) 1990-1996; 24 Access to Sanitation (% of population) 1990-1996; 25 Forest & Woodland (% of land area) 1995; 26 Annual Rate of Deforestation (%) 1990-1995; 27 CO2 Emissions per Capita (Metric tons) 1995 After the experiment with 27 variables reflecting the diverse components of life, Tony removed variables such as GDP in PPP, telephone lines, urban growth, literacy, contraceptive prevalence rate, commercial energy use, radios, and country development (dummy variable) By removing variables with high multicollinearity, he increased the significance of factors such as GDP per capita, fertility rate, school enrollment rate, and population growth whose consequence was muted due to the multicollinearity problem According to this study, GDP per capita, fertility rate, school enrollment rate, and population growth showed a high-level effect on life expectancy at birth through the regression method Gulis (2000) studied factors influencing life expectancy of 156 countries in the world He concluded that income per capita, public health spending, safe drinking water; calorie intake and literacy are the main determinants of life expectancy Hussain (2002) investigated the determinants of life expectancy by using the cross-sectional data of 91 developing countries of the world, with the help of multiple OLS through fertility rate, per capita GNP, adult literacy rate and per capita calorie intake; he studied this relationship both in linear as well as log linear model World Health Statistic (2010) suggested that Health workforce, infrastructure, and essential medicines may affect life expectancy at birth This section presents data on the resources available to the health system – this includes physicians; nurses and midwives; other health-care workers; and hospital beds These factors are essential in enabling governments to determine how best to meet the health-related needs of their populations Abdalali (2015) investigated the effects of inflation and unemployment, gross capital formation and economic development level (as economic factors), urbanity (as a social factor) and CO2 emission (as an environmental factor) on life expectancy using panel data method Through the results, the economic indices used in this study have a remarkable impact on life expectancy as well as urbanization Chhabli (2018) pointed out several socio-demographic, disease prevention indicators, lifestyle and health financing components that have some association with life expectancy Based on the result, sanitation coverage, child vaccination, and population growth are significantly associated with life expectancy at birth Amiri & Solankallio (2019) collected data from 35 OECD countries to investigate the relationship between nurse staffing and life expectancy at birth They stated that there were meaningful relationships from nurse staffing to life expectancy at birth The role of nursing characteristics in increasing life expectancy varied among different health care systems of OECD countries and on average was determined at the highest level in Japan Hence, among OECD countries, the highest effect of practicing nurses on increasing the life expectancy indicators have been investigated in Japan Rei & Takeshi (2019) made a comparison of equity preferences for life expectancy gains between Japan and Korean The study indicates that non-smokers tend to have a higher life expectancy Therefore, smoke status has a close association with life expectancy in Japan METHODOLOGY & DATA Theoretical framework of the study 1.1 Method used in research To implement this topic, our group makes hypotheses about the factors related to life expectancy at birth in Japan from 1960 to 2017 Japanese society had a lot of fluctuations since the late 1970s Because of the economic crisis, the birth rate of Japan at that time dropped significantly, affecting the average life expectancy But recently, the government has taken reforms to reverse the population situation Method of constructing econometric models: Going from studying economics, social and environmental assumptions related to the current medical issues, then construct an econometric mathematical model by defining the mathematical function form of the model with variable factors influence life expectancy at birth in Japan Method of Functional estimation: Our group use Gretl software to run model regression by using the Ordinary Least Squares method (OLS) to estimate the parameters of multivariate regression models The OLS least squares method is a simple, easy to understand and implement method; give us the optimal estimates, the properties that we want Parametric estimates by the OLS method have the following properties: Linear, Not deviate, There is the smallest variance in the class of non-biased linear estimates Method of Hypothesis test: From Gretl software we easily: Consider the differential magnification molecule VIF to identify Multi-collinearity Use the White test to test Heteroskedasticity Conducting Robust tests to identify the Autocorrelation Use the F test to evaluate the fit of the model and the t-test to estimate the confidence interval for the parameters in the model Our group proceeded to interpret the regression results, subsequently, make forecasts, analysis and finalize essays 1.2 Build theoretical models According to previous studies all over the world, in order to test the influence of factors on life expectancy, our team applied the theoretical basis as mentioned and proposed the following mathematical model: lifeexp = f (nurse, schenr, popden, comserv, frstarea, smkr) lifeexp = β1 + β2* nurse + β3* schenr + β4* popden + β5* comserv + β6* frstarea + β7* smkr + Ui In the model: ❖ Independent variables: lifeexp = Life expectancy at birth ❖ Dependent variables: nurse = Nurses and midwives schenr = School enrollment, tertiary popden = Population density comserv = Communications, computer, etc frstarea = Forest area smkr = Daily smokers in Japan 2.1 Data: The method of data collection Our team collects data on variables based on a variety of sources, which have been verified to be highly accurate The data on life expectancy at birth in Japan from 1970 - 2017 and other statistics of Nurses and midwives, School enrollment, tertiary, Population density, Communications, computer, etc - the development of technical science, Forest area were taken from the World Bank with the range from 1970 – 2017 http://microdata.worldbank.org/index.php/home Meanwhile, the data Daily smokers in Japan with the same number of years, were taken from https://ourworldindata.org/smoking https://www.cdc.gov/tobacco/data_statistics/fact_sheets/health_effects/tobacco_related_ mortality/index.htm The collected data is in the form of secondary information, collected for Japan during the period 1970 - 2017 The dataset structure is time-series and the frequency of this time-series dataset is annual 2.2 Descriptive statistics According to many theories and studies that are aforementioned in the Literature review, our group found out some economic, social and environmental factors to be variables, influence the Life expectancy at birth: Nurses and midwives, School enrollment, tertiary, Population density, Communications, computer, etc - the development of technical science, Forest area, The table below explains the reasons why those variables are chosen Table 2.2.1 Descriptive Table Independent variables Sign Name Unit Explanation of selection & Research hypothesis Life expectancy at birth reflects the overall mortality level of a population It summarizes the mortality pattern that prevails across all age groups - children Life lifeexp expectancy at birth total and adolescents, adults and the elder (years) Gains in life expectancy at birth can be attributed to a number of factors, including rising living standards, improved lifestyle and better education, as well as greater access to quality of health services Dependent variables Sign Name Unit Explanation of selection & Research hypothesis Japan is famous for the world's leading medicine and nurse Nurses and midwives per 1,000 people healthcare system The average life expectancy of the Japanese people is rated highest level on the planet Amiri & Solankallio (2019) found out the highest effect of practicing nurses on increasing the life expectancy indicators Increase the number of doctors, Nurses and midwives, Improve health services at health facilities, factors that can 10 improve the quality of health services, which make the life expectancy at birth in Japan increase Japan aims to ensure the harmonious development of children in all aspects: heart, intellect, affection, spirit, attitude, value system, humanity Japan is one of the developed countries in the world with the actual illiteracy rate of and 72.5% of the students schenr School enrollment, % gross tertiary attending up to university, college and secondary The development of education is the basis to ensure stable social development and quality of life Tony Smith (2000) also showed a high-level effect of School enrollment, tertiary on life expectancy at birth through the regression method The higher % School enrollment, tertiary is, the higher life expectancy at birth in Japan reaches Japan's population is 127.2 million, accounts for th hundred popden Population density people per sq km of land area 1.68% of the world's population, ranks 11 in the topmost populous country in the world However, Japan is facing the situation to be felt into the aging labor force in the world The relation between Population density and life expectancy has been pointed out by Tony Smith (2000) The growth in the Japanese life expectancy is the result of the increase in population density For a long time, Japanese technology has surprised comserv Communic ations, computer, etc ‰ of service imports, BoP the whole world by the speed of development and cutting-edge innovations, which are favorable conditions for knowledge development and national medical system It’s obviously that the Japanese population made great strides in achieving higher 11 schenr 0.18410 0.65065 0.27300 popden 2.9010 3.5130 0.27788 comserv 3.1083 6.2580 0.75189 frstarea 6.8273 6.8478 0.0065131 smkr 22.348 33.007 4.8129 (Source: the team synthesized under the support of Gretl software) From the table, we can easily see that: ❖ ❖ ❖ Life expectancy at birth of Japanese from 1960 to 2017 fluctuates from 71.95 to 84.100, the average value is 79.415, the mean is 79.275 Numbers of Nurses and midwives in Japan per 1,000 people from 1960 to 2017 fluctuates from 5.567 to 11.638, the average value is 7.0945, the mean is 7.943 School enrollment, tertiary index (% gross) reflect quality of education from 1960 to 2017 fluctuates from 0.17200 to 0.65800, the average value is 0.35750, the mean is 0.40698 ❖ Population density (hundred people per sq km of land area in Japan) from 1960 to 2017 fluctuates from 2.8455 to 3.5134, the average value is 3.4215, the mean is 3.3409 ❖ ‰ of service imports, BoP, Communications, computer, etc show the development of technical science from 1960 to 2017 fluctuates from 3.0171 to 6.3172, the average value is 4.2975, the mean is 4.3630 ❖ ‰ of land area covered with Forest in Japan from 1960 to 2017 fluctuates from 6.8247 to 6.8484, the average value is 6.8387, the mean is 6.8390 Number of Daily smokers in Japan from 1960 to 2017 fluctuates from 21.734 to 33.059, the average value is 31.757, the mean is 29.809 ❖ 14 Table 2.2.3: Correlation coefficients, using the observations 1970 - 2017 5% critical value (two-tailed) = 0.2845 for n = 48 lifeexp nurse schenr popden comserv 1.0000 0.9177 0.9487 0.9474 0.8477 lifeexp 1.0000 0.9836 0.7618 0.8407 nurse 1.0000 0.8199 0.8332 schenr 1.0000 0.7286 popden 1.0000 comserv frstarea smkr 0.2351 -0.6784 lifeexp 0.3964 -0.8661 nurse 0.3050 -0.7996 schenr 0.0387 -0.4285 popden 0.3901 -0.7505 comserv 1.0000 -0.7447 frstarea 1.0000 smkr (Source: the team synthesized under the support of Gretl software) From the table, we can easily see that: ❖ Correlation coefficients between dependent variable Life expectancy and Nurses and midwives, School enrollment, tertiary, Population density is very high The Correlation coefficients are 0.9177, 0.9487 and 0.9474 respectively, much > 0.8 The correlation relationship is up hill It shows that these factors Nurses and midwives, 15 School enrollment, tertiary, Population density has strong impacts on Life expectancy at birth index in Japan ❖ Correlation coefficients between dependent variable Life expectancy and Daily smokers in Japan is -0.6784 The correlation relationship is down hill, which shows the dimensional backward effects between the Life expectancy and number of daily smokers ❖ Besides, Correlation coefficient between independent variables Nurses and midwives and School enrollment, tertiary is also very high 0.9836 The correlation relationship is up hill It shows that number of Nurses and midwives has much relevant to School enrollment, tertiary index ❖ Almost correlation coefficients are positive, showing the positive relationship between the dependent and independent variables Except that correlation coefficients between Daily smokers in Japan and the rest of independent variables Nurses and midwives, School enrollment, tertiary, Population density, Forest area are negative, - 0.8661, -0.7996, -0.4285, -0.7505, -0.7447 in the order given ❖ Similarly, correlation coefficient between Forest area and other variables not too high, fluctuate moderately from 0.2351 to 0.3964 From there, it can see that the expectation sign is a positive sign ("+") 16 ESTIMATED MODEL I Estimated results: Using Gretl software as well as Ordinary Least Squares Method, we get the following results: Table I.1: Summary Table Variables Model Model Const 340.064*** 340.064*** (3.203) (2.969) −0.292884** −0.292884** (−2.057) (−2.380) 6.46339*** 6.46339*** (4.096) (3.933) 11.2830*** 11.2830 (21.55) (21.68) 0.175457** 0.175457 (2.021) (1.630) −42.4916*** −42.4916** (−2.767) (−2.565) −0.300395*** −0.300395*** (−5.112) (−4.958) Nurse Schenr Popden Comserv Frstarea Smkr 17 N R 48 48 0.996167 0.996167 Heteroscadasticity White's test for heteroskedasticity - Auto-colleration White's test for heteroskedasticity - Null hypothesis: heteroskedasticity not present Null hypothesis: heteroskedasticity not present Test statistic: LM = 24.7235 Test statistic: LM = 24.7235 with p-value = P (Chi-square (27) > 24.7235) = 0.589952 with p-value = P (Chi-square (27) > 24.7235) = 0.589952 LM test for autocorrelation up to order - LM test for autocorrelation up to order - Null hypothesis: no autocorrelation Null hypothesis: no autocorrelation Test statistic: LMF = 7.4122 Test statistic: LMF = 7.4122 with p-value = P (F (1, 40) > 7.4122) = 0.00955118 with p-value = P (F (1, 40) > 7.4122) = 0.00955118 Note: * means that the coefficients are significant at 1% level * means that the coefficients are significant at 5% level (number): t-ratio 18 II Data explanation ❖ Number of observations: n = 48 ❖ Mean dependent var = 79.27510 ❖ Sum squared resid = 1.941228 ❖ ̂ ❖ = 0.217594 R-squared = 0.996167 shows that independent variables can explain 99.61% the change of value of the life expectancy-dependent variable, the rest is contributed by other factors That R2 = 0.996167 is quite high, which suggests that the model is good fit ❖ Adjusted R-squared = 0.995607 Meaning of coeffiecients in the regression model: ▪ ▪ ▪ ▪ ▪ ▪ = −0.292884 When nurse increases by (per 1,000 people), holding the value of the other variables constant, the estimated value of lifeexp decreases by 29,29 (years) = 6.46339 When schenr increases by (% gross), ceteris paribus, the estimated value value of lifeexp increases by 646.34 (years) = 11.2830 When popden increases by (hundred people per sq km of land area), holding the value of the other variables constant, the estimated value of lifeexp increases by 1128.3 (years) = 0.175457 When comsevr increases by (‰ of service imports, BoP), ceteris paribus, the estimated value value of lifeexp increases by 17.55 (years) = −42.4916 When frstarea increases by (‰ of land area), holding the value of the other variables constant, the estimated value of lifeexp decreases by 4249.16 (years) = −0.300395 When smkr increases by (million people), ceteris paribus, the estimated value value of lifeexp decreases by 30.04 (years) From Model 1, we have Population Regression Function as below: = − ∗+ ∗+ ∗+ ∗− ∗− ∗ ̂ 19 20 III Testing: Heteroskedasticity Heteroskedasticity indicates that the variance of the error term is not constant, which makes the least squares results no longer efficient and t tests and F tests results may be misleading The problem of Heteroskedasticity can be detected by plotting the residuals against each of the regressors, most popularly the White’s test It can be remedied by respecifying the model – look for other missing variables Let us discuss by using White Test detect the presence of heteroskedasticity Given that the hypothesis is: { 0: ℎ ℎ 1: ℎ ℎ Run White’s Test Command in Gretl, we have: Table III.1: Result of White Test for Heteroskedasticity White's test for heteroskedasticity Null hypothesis: heteroskedasticity not present Test statistic: LM = 24.7235 with p-value = P (Chi-square (27) > 24.7235) = 0.589952 As we can see that p-value = 0.589952 > 0.05 At the 5% significance level, there is enough evidence not to reject the null hypothesis and conclude that this set of data does not meet the problem of Heteroskedasticity 21 Multicollinearity: Multicollinearity is the high degree of correlation amongst the explanatory variables, which may make it difficult to separate out the effects of the individual regressors, standard errors may be overestimated and t-value depressed The problem of Multicollinearity can be detected by examining the correlation matrix of regressors and carry out auxiliary regressions amongst them In Gretl, the VIF command is used, which stand for variance inflation factor As we can infer from the Summary Table that there is at least a t-ratio of which the value is bigger than 2, indicating that Multicollinearity is not too worrisome a problem for this set of data Autocorrelation Autocorrelation is where error terms in a time series transfer from one period to another In other words, the error for one time period a is correlated with the error for a subsequent time period b For example, an underestimate for one quarter’s profits can result in an underestimate of profits for subsequent quarters Given that the hypothesis is: { : ℎ ℎ 1: ℎ Run White’s Test Command in Gretl, we have: LM test for autocorrelation up to order Null hypothesis: no autocorrelation Test statistic: LMF = 7.4122 with p-value = P (F (1, 40) > 7.4122) = 0.00955118 Table III.3: Result of Auto Correlation Test As we can see that p-value = 0.00955118 < 0.05 At the 5% significance level, there is enough evidence to reject the null hypothesis and conclude that this set of data meets the problem of Autocorrelation Problem 22 To fix the problem, robust standard errors are used to relax the assumption that errors are both independent and identically distributed In Gretl, regression is rerun with the robust option, using the command: Robust Standard Error Model 2: OLS, using observations 1970-2017 (T = 48) Dependent variable: Lifeexp HAC standard errors ticked on Model 2: OLS, using observations 1970-2017 (T = 48) Dependent variable: lifeexp HAC standard errors, bandwidth (Bartlett kernel) T Const Nurse Schenr Popden Comserv Frstarea Smkr Coefficient 340.064 −0.292884 6.46339 11.2830 0.175457 −42.4916 −0.300395 Mean dependent var Sum squared resid R-squared F(6, 41) Log-likelihood Schwarz criterion Rho Std Error 114.550 0.123059 1.64347 0.520530 0.107667 16.5633 0.0605864 79.27510 1.941228 0.996167 1346.558 8.880075 9.338258 0.391171 S.D dependent var S.E of regression Adjusted R-squared P-value(F) Akaike criterion Hannan-Quinn Durbin-Watson LM test for autocorrelation up to order Null hypothesis: no autocorrelation Test statistic: LMF = 7.4122 with p-value = P F (1, 40) > 7.4122) = 0.00955118 23 t-ratio 2.969 −2.380 3.933 21.68 1.630 −2.565 −4.958 p-value 0.0050 0.0220 0.0003 7.50409) = 0.00616 Ljung-Box Q' = 7.30163, with p-value = P(Chi-square(1) > 7.30163) = 0.00689  The Model is still Autocorrelated but it is much improved 24 CONCLUSION Life expectancy is a comprehensive index of the health status in a community or a country and useful for comparing health conditions as a whole Although Japan is famous for having the highest life expectancy in the world, there has not been any up-to-date papers analyzing factors influencing life expectancy at birth in Japan Therefore, our group decided to study the indicators affecting life expectancy at birth in Japan from 1970-2017 In our study, we used Life Expectancy at Birth as the dependent variable while Nurses and midwives, School enrollment, tertiary, Population density, Communications, computer, etc - the development of technical science, Forest area and Daily smokers are regressed as the independent variables in our OLS regression model on Gretl Our initial model has auto-correlation problem and we managed to solve it by using the Robust standard error to get the results as interpreted below The estimated results indicate that School enrollment and Daily smokers are statistically significant at 1% level, which means they have remarkably high effects on Life Expectancy at Birth in Japan School enrollment and Life Expectancy at Birth in Japan have a noticeably positive relationship, meanwhile, that of Daily smokers and Life Expectancy at Birth is negative These statistics approved our research hypothesis in the Methodology & Data session The two independent variables, Nurses and midwives and Forest area are statistically significant at 5% level Both of them display a negative relationship with Life Expectancy at Birth, which means that an increase in the number of nursing staff and in forest area leads to a decrease of life expectancy at birth This finding goes against our research hypothesis initially, however, a study of the related literature disclosed that there is a possibility of bias, called ecological fallacy in the analysis of correlation at the population level Hence, further research needs to be done to fulfill the constraints of this study and to provide a reasonable explanation for the phenomenon It can be seen that although the models have a defect, the independent variables have great influence on the dependent variable, we hope to more related research, find more documents to overcome and transform them into better models As a result, other developed countries with the same level of economic development can learn lessons from the indicators affecting life expectancy of Japan to leverage the 25 lifespan of a human, especially through encouraging school enrollment and limiting the number of daily smokers Due to the lack of time and source of data, there are remaining critical factors that we could not include in our regression model Thus, our study still has some space for improvement We would love to receive constructive comments on our project to develop our further study We hope that in the future, this topic will still be an inspiration for researchers to investigate Finally, we would love to send appreciation towards our instructors Dr Tu Thuy Anh and MSc Chu Thi Mai Phuong for helping us throughout the process of implementing our project REFERENCES Ali Amjad, A K (2014) The Impact of Socio-Economic Factors on Life Expectancy for Sultanate of Oman: An Empirical Analysis Andreas E Stucka, J M (1999, February) Risk factors for functional status decline in community-living elderly people: a systematic literature review Social Science & Medicine, 48, 445-469 Ardo van den Hout, E O (2014, April 15) Investigating healthy life expectancy using a multi-state model in the presence of missing data and misclassification Demographic Research, 30, 1219-1244 Arshia Amiri, T S.-V (2019) Nurse staffing and life expectancy at birth and at 65 years old: Evidence from 35 International Journal of Nursing Sciences Asakura, D M (1989, April) Ageing and health in Japan Journal of CrossCultural Gerontology, 4, 143–162 Barro, R J (1996) Determinants of Economic Growth NBER Working Paper Series (5698) Chhabi L Ranabhat, J A.-B.-B (2018, September 18) The Influence of Universal Health Coverage on Life Expectancy at Birth (LEAB) and Healthy Life Expectancy (HALE): A Multi-Country Cross-Sectional Study Front Pharmacol 26 Donald Willcox, S Y (2012, January) Gender gap in healthspan and life expectancy in Okinawa: Health behaviours Asian J Gerontol Geriatr 2012, 7, 49– 58 Dorly J.H., D T (2002, April) Attrition in the Longitudinal Aging Study Amsterdam: The effect of differential inclusion in side studies 55, 319-328 Dr Colin D Mathers, R S (2001, May 8) Healthy life expectancy in 191 countries, 1999 The Lancet, 357, 1685-1691 Jessica Y Ho, A S (2018, August 15) Recent trends in life expectancy across high income countries: retrospective observational study BMJ Jim Oeppen, J W (2002, May 10) Broken Limits to Life Expectancy Policy Forum, 296, 1029-1031 Jong In Kim, G K (2016, October) Country-Level Socioeconomic Indicators Associated with Healthy Life Expectancy: Income, Urbanization, Schooling, and Internet Users: 2000–2012 Social Indicators Research, 129, 391–402 Juan Suna, P S (2004, July) The impact of uterine cancer on life expectancy in Japan International Journal of Gynecology & Obstetrics, 86(1), 54-55 KazushiOkamoto (2006, July-August) Life expectancy at the age of 65 years and environmental factors: An ecological study in Japan Archives of Gerontology and Geriatrics, 43(1), 85-91 Lotfali Agheli, S E (2016, February) Determinants of Life Expectancy at Birth in Iran: A modified Grossman Health European Online Journal of Natural and Social Sciences, 4, 427-437 Mark D Chatfielda, C E (2005, January) A systematic literature review of attrition between waves in longitudinal studies in the elderly shows a consistent pattern of dropout between differing studies Journal of Clinical Epidemiology, 58, 13-19 Momoko Iwamotoa, F N (2014, October) Estimated life expectancy and risk of death from cancer by quartiles in the older Japanese population: 2010 vital statistics Cancer Epidemiology, 38, 511-514 Motofumi Masaki, A K (1987, February) Increase in life expectancy at birth in Japan: Some implications for variable patterns of decrease in mortality Health Policy, 7, 41-48 27 Keita Konno, Yuriko Katsumata, Asuna Arai, Hiko Tamashiro, A l (2004, March-April) Functional status and active life expectancy among senior citizens in a small town in Japan Archives of Gerontology and Geriatrics, 38(2), 153-166 Poulain, M (2011, July 21) Exceptional longevity in Okinawa: A plea for indepth validation Demogaphic Research, 25, 245-284 Rei Goto, T M (2019, May) Comparison of Equity Preferences for Life Expectancy Gains: A Discrete Choice Experiment Among the Japanese and Korean General Public Value in Health Regional Issues, 18, 8-13 Ruhai Bai, J W (2018, December 7) Trends in Life Expectancy and Its Association with Economic Factors in the Belt and Road Countries—Evidence from 2000–2014 International Journal Environment Research and Public Health, 15 Sabine K Schnabel, P H (2009, August 7) An analysis of life expectancy and economic Demographic Research, 21, 109-134 Smith, T (2000) A Statiscal Analysis of Life Expectancy across Countries using Multiple Regression Statistics, W H (2010) Global Health Indicators Tanja AJ Houweling, A E (2001, May) World Health Report 2000: inequality index and socioeconomic inequalities in mortality The Lancet, 357, 1671-1672 Upreti, P (2017) Factors Affecting Economic Growth in Developing Major Themes in Economics , 17 Vahid Yazdi Feyzabi, A H (2015) The Association between Peace and Life Expectancy: An Empirical Study of the World Countries Iran J Public Health, 44, 341–351 Vanessa Yong, Y S (2009, April 28) Trends in healthy life expectancy in Japan: 1986 – 2004 Demographic Research, 20, 467-494 Wilkinson, R G (1992) Income distribution and life expectancy Xian Liu, J L (1995) Transitions in Functional Status and Active Life Expectancy among Older People in Japan The Journals of Gerontology, 50B, S383–S394 Y.Murakamia, R T (2018, July) Comparison of methods for calculating healthy life expectancy in a Japanese population: An analysis of Nippon Data90 Revue d'Épidémiologie et de Santé Publique, 66, S329-S330 28 ... highly urgent topic ? ?Factors influencing life expectancy at birth in Japan from 1960 to 2017? ?? The objective of our assignment is to investigate the indicators and make recommendations, suggestions... whole Although Japan is famous for having the highest life expectancy in the world, there has not been any up-to-date papers analyzing factors influencing life expectancy at birth in Japan Therefore,... Therefore, our group decided to study the indicators affecting life expectancy at birth in Japan from 1970- 2017 In our study, we used Life Expectancy at Birth as the dependent variable while Nurses