Page 122 - GHES-2-1

Global Health Econ Sustain Prolonged impact of health-care expenditure on poverty

impact of HEt on Pt. The error term (ε1) accounts for slow adjustment speed toward equilibrium, signifying that
factors not included in the model. deviations from the equilibrium level will take almost
Pt=1.600764+1.921485⋅HE t−1 (XII) five years to be rectified. Moreover, the short-run ARDL
model, with significant outcomes at a 5% level according to
3.4. Diagnostic tests p-values, yields an adjusted R-squared of 71%, suggesting
Diagnostic tests applied to the model, the Breusch-Pagan- that the model adequately represents the population.
Godfrey heteroskedasticity test for the model results in Moreover, the F-statistic signifies the model’s predictive
Table 6, imply that since the probability value of F-statistics ability for the dependent variable, the poverty rate.
and Obs*R-squared is >5%, the null is rejected and there The findings highlight a noteworthy positive
is no heteroskedasticity. Meanwhile, the Ramsey RESET association between healthcare expenditure and poverty
Test result indicates that the data have been stationary rates. The equation derived from the model indicates
because the t- and F-statistical probability values are >5%, that a unit increase in health-care expenditure leads to
and the model is linear. Moreover, the Breusch-Godfrey a 1.92 unit increase in the poverty rate in the long run,
serial correlation LM test explains that the F-statistic and all else constant. This conclusion suggests that enhancing
Obs*R-squared probability results are >5%, which implies healthcare spending in low- and middle-income
that there is no serial correlation between the model countries may inadvertently contribute to an increase in
variables. poverty levels.
In addition, the cointegration test results in Table 7 4.1. Causality, caveats, and generalizability
show that since the probability is 0.0000, which is <5%, we
have cointegration, which is good for long-run equation Considering the aftermath of the global pandemic, it
answering. Finally, the stability of the test is evident is essential to take into account the complex effects that
using the CUSUM test, where the cumulative sums of fiscal and monetary policies have on emerging economies.
deviations from a specified reference value are computed According to Cortes et al. (2022) and Benmelech & Tzur-
over time. This involves calculating the cumulative sum Ilan (2020), the financial assistance provided to people and
of the differences between the observed and expected or enterprises results in a complicated economic environment.
predicted values. As depicted in Figure 2, the CUSUM Providing such assistance can indeed address immediate
plot exhibited random fluctuations around zero, indicating financial challenges, but it carries the risk of disrupting
significant structural changes, which suggests that there capital flows and influencing asset values, which, in turn,
are no issues with recursive residuals in terms of the may impact the economic factors under examination.
mean. This was because the fluctuations were within a 5% For example, policies implemented in response to the
significance range. pandemic is highlighted the multiple difficulties, Desson
et al. (2020). Inflation tends to elevate poverty levels and
4. Discussion exacerbate economic disparities. In addition, as Kose et al.

The empirical analysis conducted using the ECM and ARDL (2022) pointed out, the effect on the actual value of debt
models suggests a crucial short- and long-term relationship changes the fiscal space available to emerging economies.
between healthcare expenditure and the poverty rate. The In addition, as Kose et al. (2022) pointed out, the effect on
negative and statistically significant estimated coefficient the actual value of debt changes the fiscal space available to
of the ECT in the ECM model (-0.183745) indicates a emerging economies. In addition, Campello et al. (2020)
long-run equilibrium relationship between poverty and noted that the pandemic’s impact on business recruiting
healthcare spending. The ECT of 18.37% implies a relatively adds to changes in household income and patterns of

Table 6. Diagnostic test
Diagnostic test Statistics Column 1 Column 2 P‑value
Serial correlation test F-statistic 1.395681 Prob. F (1,373) 0.2382
Obs*R 2 1.409118 Prob. Chi-square (1) 0.2352
Heteroskedasticity test F-statistic 0.055834 Prob. F (4,373) 0.9942
Obs*R 2 0.226195 Prob. Chi-square (4) 0.9941
Ramsey RESET test t-statistic 1.371283 377 0.1711
F-statistic 1.880417 (1, 377) 0.1711
Abbreviation: Prob.: Probability

Volume 2 Issue 1 (2024) 6 https://doi.org/10.36922/ghes.2383