Does financial literacy influence preventive health checkup behavior in Japan? a cross-sectional study | BMC Public Health

Data

We used Osaka University’s Preference Parameters Study (PPS) for 2010 and 2011. The PPS was a panel survey conducted annually from 2003 to 2013 that extracted information on socioeconomic and demographic characteristics and Japanese preferences. We extracted data on financial literacy as the main independent variable and on financial education as the second significant primary independent variable, from the 2010 dataset. From 2011, we obtained behavioral data of the health check as the dependent variable, as well as data on the control variables. Our sample focused on the middle-aged Japanese working population (40-64 years old), for whom health check-ups were considered more important. In addition, health check-ups and advice, focusing on visceral obesity, target policyholders and dependents aged 40 or over. [34]. We combined the datasets using each respondent’s panel credentials; 3,420 observations remained after excluding samples with missing demographic and socioeconomic variables and unmatched data. Removing 1,212 individuals aged under 40 and over 64 yields the final sample of 2,208 responses, which represents 44.75% of the total responses in the 2011 dataset (4 934 sightings).

variables

A complete overview of variable definitions is provided in this subsection. To begin with, the dependent variable indicates the probability of having a health check; that is, the odds that an individual has had at least 1 type of health check-up per year, excluding health check-ups organized by employers and schools. One of the questions in the PPS 2011 dataset asks about each respondent’s behavior regarding health checks: “(In the past 12 months) Have you had health checks (at the exclusion of cancer examination, prenatal check-up, and medical treatment)”, with six possible answers: “1 – Health check-up organized by the municipality”, “2 – Health check-up organized by your employee or the your employee”, “3 – Health check-up organized by your school”, “4 – Medical check-up (other than or greater than 1–3)”, “5 – Other”, “6 – I did not have a check-up From these answers, we created the dependent variable as a binary variable equal to 1, for any type of health check done in the year apart from those organized by the government, employers and schools, because these check-ups are compulsory, so the idea of ​​rationality does not apply.

We had two main variables of interest: financial literacy and financial education, which assess financial knowledge from the perspective of investing and saving behaviors, respectively. The first is based on the three questions developed by Lusardi and Mitchell [35] provided in the appendix. These questions test an individual’s mathematical abilities and understanding of basic financial concepts such as interest rates, inflation, and risk diversification, which are essential for making sound investments. Because of their simplicity and empirical support, several other studies have adopted these questions to measure financial literacy. [3, 6, 36,37,38,39]. To quantify the financial literacy questions, we assigned a score of 1 for each correct answer and 0 for an incorrect answer. We took the equally weighted average scores of the three questions to obtain the financial literacy variable.

The other main explanatory variable—financial education—was based on the following multiple-choice question from the PPS survey: “Did you receive compulsory financial education when you were in elementary school?” We assigned a value of 1 if the respondent answered ‘yes’ and 0 if they answered ‘no’ or ‘don’t know’. We included this option because Japan is the only country in Asia to offer financial education programs in its school curriculum. [40]. These programs potentially influence financial stability risks and the corresponding long-term preventive demand. In addition, financial education has no effect on the financial literacy of Japanese people who have already received this type of education in the program. [41]; therefore, we evaluate both variables.

Other demographic and socioeconomic variables included gender, age, university degree, marital status, number of household members, children, employment status, household income, and household assets. We also controlled for health risk behaviors such as smoking, alcohol consumption and gambling addiction. Finally, we included psychological variables such as myopic view of the future, current level of happiness , anxiety about health and poor health. A summarized version of the variables used, their respective types, definitions and codings is provided in Table 1.

Table 1 Variable definitions

Methods

As our dependent variable was a binary variable, we performed a probit regression analysis. This was done to test our hypothesis that people with better financial literacy and better financial education are more likely to participate in non-mandatory preventative health checkups. Equations 1 and 2 assess the impacts of financial literacy and financial education on health check behavior, individually. Model (3) incorporates the two variables to assess their combined impact.

$${{varvec{Y}}}_{{varvec{i}}} = {varvec{f}}({{varvec{F}}{varvec{L}}}_{{ varvec{i}}}, {{varvec{X}}}_{{varvec{i}}},{{varvec{varepsilon}}}_{{varvec{i}}}),$ $

(1)

$${{varvec{Y}}}_{{varvec{i}}}boldsymbol{ }=boldsymbol{ }{varvec{f}}({{varvec{F}}{varvec{ E}}}_{{varvec{i}}},boldsymbol{ }{{varvec{X}}}_{{varvec{i}}},{{varvec{varepsilon}}}_ {{varvec{i}}}),$$

(2)

$${{varvec{Y}}}_{{varvec{i}}}boldsymbol{ }=boldsymbol{ }{varvec{f}}({{varvec{F}}{varvec{ L}}}_{{varvec{i}}},boldsymbol{ }{{varvec{F}}{varvec{E}}}_{{varvec{i}}},boldsymbol{ } {{varvec{X}}}_{{varvec{i}}},{{varvec{varepsilon}}}_{{varvec{i}}}),$$

(3)

where ({Y}_{i}) indicates the preventative health check behavior of the (I) and answering, (FL) represents the average financial literacy score, (FE) is the financial education status, (X) is a vector of the characteristics of the respondents, and (varepsilon) is the error term. We used a probit model to predict health checkup behavior with respect to financial literacy and financial education, after adjusting for socioeconomic and demographic characteristics. This model describes the probability that an event falls into one of the specified categories.

To identify and correct for high intercorrelations between two or more independent variables in all models, we performed correlation and multicollinearity tests (available upon request). For example, people with a high level of education may have high financial knowledge, or those with a high net worth may have more financial knowledge due to their experience in asset management. The correlation matrix revealed a weak relationship between the relative movements of two variables in all models (significantly less than 0.70). Moreover, the tests of the inflation factor of the variance of the explanatory variables showed a non-significant presence of multicollinearity in all the models.

Although potential endogeneity is a concern, we believe this will not affect our results, for several reasons. First, we included important factors that may influence the health check behavior of individuals. These include socio-economic, demographic and health characteristics of individuals which are widely used in various health economics studies to explain phenomena. Second, the theoretical background discussed in the introductory sections establishes the basis for how financial literacy might influence health check behavior and why reverse causation is less likely to occur. Finally, having a one-year lag in financial literacy scores relative to health check behavior and other data also minimizes the likelihood of reverse causation.

We have created four templates for the equations. (1–3), each with a separate control variable. We have provided an example of our model requirements for Eq. (1) below. Equations (4), (5), (6) and (7) represent models 1, 2, 3 and 4, respectively.

$$Probability of being healthy {balance sheet}_{i} = Phi left( {beta }_{0}+{beta }_{1}{financial literacy}_{i}+{ beta }_ {2}{man}_{i}+{beta }_{3}{age}_{i}+ {beta }_{4}university {diploma}_{i}+{beta }_{ 5}{marriage}_{i}+{beta }_{6}{divorce}_{i}+{beta }_{7}{household size}_{i}+{ beta }_{8 } {children}_{i}+{beta }_{9}{unemployed}_{i}+{beta }_{10}{log household income}_{i}+ {beta }_{11 }{household goods logo}_{i}right),$$

(4)

$${Probability of getting a health check}_{i}=Phi ({ beta }_{0}+{ beta }_{1}{financial literacy}_{i}+{beta } _{2 }{male}_{i}+{beta }_{3}{age}_{i}+{beta }_{4}{university degree}_{i}+{beta }_ {5} {marriage}_{i}+{beta }_{6} {divorce}_{i}+{beta }_{7}{household size}_{i}+ {beta }_ {8}{ children}_{i}+{beta }_{9} {unemployed}_{i}+{ beta }_{10}{log household income}_{i}+{ beta }_{11}{ household goods logo}_{i}+{beta }_{12}{current smoker}_{i}+{beta }_{13}{current drinker}_{i }+ {beta }_{14 }{frequent players}_{i} ),$$

(5)

$${Probability of getting a health check}_{i}= Phi ({ beta }_{0}+{beta }_{1}{financial literacy}_{i}+{beta } _{2 }{male}_{i}+{beta }_{3}{age}_{i} + {beta }_{4}{university degree}_{i}+{beta }_ {5} {marriage}_{i}+{ beta }_{6}{divorce}_{i}+{beta }_{7}{household size}_{i}+ {beta }_ {8}{ children}_{i}+{beta }_{9}{unemployed}_{i}+{ beta }_{10}{log household income}_{i}+{ beta }_{11}{household goods logo}_{i}+{beta }_{12}{current smoker}_{i}+{beta }_{13}{current drinker}_{i } + {beta }_{14 }{frequent players}_{i}+{ beta }_{15}{myopic view of the future}_{i} ),$$

(6)

$${Probability of getting a health check}_{i}= Phi ( {beta }_{0}+{ beta }_{1}{financial literacy}_{i} +{beta } _{2 }{male}_{i} +{beta }_{3}{age}_{i} +{beta }_{4}{university degree}_{i} +{beta }_ {5} {marriage}_{i} + {beta }_{6}{divorce}_{i} +{beta }_{7}{household size}_{i}+{ beta }_ {8}{ children}_{i} +{beta }_{9}{unemployed}_{i} + {beta }_{10}{log household income}_{i} +{ beta }_{11}{ household goods logo}_{i} +{beta }_{12}{current smoker}_{i} +{beta }_{13}{current drinker}_{i } + {beta }_{14 }{frequent players}_{i} + {beta }_{15}{myopic view of the future}_{i} +{beta }_{16}{level happiness}_{i} + { beta }_{17}{anxiety about health}_{i} +{beta }_{18}{poor health}_{i}). $$

(seven)

Sarah J. Greer