Technical Notes of COVID-19 Research

Methodologies and construction of indices used in the comparative research on governmental actions during the pandemic

Technical Notes

The technical notes summarize the QCA methodology and the one used to construct the indices created within the scope of the research: the Government Actions Index (IAG), which is the phenomenon to be explained, and the indices referring to the explanatory factors of the IAG: Welfare Index (IBE), Health System Index (ISS) and Adherence to WHO Recommendations. For further details, see the report available on this site. For information on the two other indices that were used as explanatory factors, see: for the HDI: the HDI calculation methodology undergoes periodic changes; for more information see the Human Development Reports published annually by UNDP. For the DEM: https://www.economist.com/interactive/democracy-index.

Constructed Indices

Four main indices were developed to capture the central dimensions of the research:

IAG

Classification of governmental actions in response to the pandemic

Sources: OxCGRT Tracker, IMF Fiscal Monitor

Period: 2020-2021

Scale: 0 to 1 (average RG, PV, GARF Health)

Click to see details →

ISS

Classification of health systems by their effective implementation

Dimensions: Public Investment, Public Expenditure, Quality

Period: 2015-2019

Scale: 0 to 1 (effective implementation)

Click to see details →

IBS

State capacity to provide social protection

Method: Principal Component Analysis

Base: 11 studies, 13 classifications

Scale: 0 to 1 (state social protection)

Click to see details →

QCA

Analytical method for configurational comparative studies

Type: Configurational method

Function: Multicausal analysis

Application: Fuzzy sets

Click to see details →

IAG - Government Action Index

Classification of governmental actions in response to the pandemic

IAG Construction Method

To classify and rank the governmental actions of countries in confronting the pandemic crisis, data from the OxCGRT Tracker and IMF Fiscal Monitor were used so that the information would be comparable, consistent, and cover the peak period of the pandemic, 2020 and 2021.

The IAG anchors the classification of each country in the RG Index - formed by the Containment (C), Health (S) and Economic Support (SE) indices - and the Vaccine Policies Index (PV), all from the OxCGRT Tracker, as well as additional spending or tax waiver by national governments in health per capita (GARF Health) from the IMF Fiscal Monitor.

The IAG is an index from 0 to 1 calculated by the average of the RG Index, PV Index and GARF Health per capita, each calculated as indices from 0 to 1.

The higher the IAG value, the more complete the government actions adopted in that country were. The index was developed to compare the governmental actions of the 14 countries analyzed. Thus, the value '0' does not indicate that the national government of the country did not implement actions and the value '1' does not mean that the government was perfect in its actions, but only that it is at the top in relation to the other cases selected in the sample.

ISS - Health Systems Index

Classification of health systems by their effective implementation

ISS Construction Method

The dimensions considered to classify and rank health systems do not refer to the formal characteristics of these systems, but to their effective implementation, based on the assumption that a more inclusive system, or one with a greater relative degree of coverage through public financing, with greater public spending and better quality, would be more favorable for a more effective pandemic response. The index is formed from three dimensions and several indicators, as summarized in the table below:

Dimensions, indicators, measurement and sources

Dimensions	Indicators	Measurement and classification	Sources
Degree of public investment in health vis-à-vis household spending (2019).	Combination of: i) % of public health spending relative to total health spending; and ii) % of out-of-pocket spending relative to total health spending.	i) very restricted relative public investment: public spending less than 60% and out-of-pocket greater than 25%; ii) restricted relative public investment: public spending greater than 60% and out-of-pocket greater than 25%; iii) high relative public investment: public spending greater than 60% and out-of-pocket less than 25%.	World Bank (2024).
General government domestic health expenditure (per capita, PPP) (2019).	Public health expenditure per capita in US$.	Index values range from 0–1, with 1 being the best value.	World Bank (2024).
Health care quality (average for 2015–2019).	i) Maternal mortality per 100,000 live births; ii) Neonatal mortality per 1,000 live births; iii) Vaccine coverage (7 types of vaccines); iv) Proportion of deaths due to ill-defined and ignored causes; v) Prenatal care coverage (minimum of 4 visits).	Indices are calculated for each indicator and then aggregated as the average of the five indicators.	WHO (2024).

IBS - Social Welfare Index

State capacity to provide social protection

IBS Construction Method

The Social Welfare Index (IBS) is a synthetic indicator that classifies 14 of the 20 Latin American countries regarding state capacity to provide social protection to their citizens before the COVID-19 pandemic. The technique used for its construction is Principal Component Analysis (PCA).

For the analysis, 18 studies on Latin American welfare regimes were selected that provide country classifications in different typologies. These studies vary according to the empirical universe and period analyzed and the methodology mobilized for comparing provisions guaranteed by the State, exclusively or together with those from the market and families.

For the construction of the index, only 11 studies were selected that focus on provisions guaranteed by the State, that is, social policies, and propose 13 country classifications. The argument that supports the construction is that the basic classification criterion - the performance of countries in quantitative indicators - allows capturing, through the technique mobilized, the consensus built in the area about the nature of state social protection guaranteed in the selected Latin American countries.

The construction of the IBS was done in four stages:

1mapping the classification given to countries in different typologies;
2translation of types into ordinal numerical values, according to the performance denoted in the categorical classification;
3performing the tests and analyses required by the PCA technique;
4index calculation.

The IBS should be interpreted as follows: the scale ranges from 0 to 1, and the higher its value, the greater the social protection offered by the State in that country. Another important point is to note that the index indicates the relative position of each of the 14 countries selected for research in the ranking, and serves only this purpose.

QCA - Qualitative Comparative Analysis

Analytical method for configurational comparative studies

QCA Methodology

The QCA methodology follows these steps:

Calibration: transform indicators (original values) into fuzzy sets (values from 0 to 1), making them operational for QCA. This calibration must set upper, lower and intermediate thresholds. The upper threshold indicates full membership in the set (value 1.00), the lower threshold indicates full non-membership (0.00) and the intermediate threshold receives 0.50. In the case of the IAG, the intermediate threshold was set at 0.6.
Analysis of necessary conditions. A condition is considered necessary when whenever Y (outcome) is present, X (condition) is present (X≤Y). To identify necessity, we used the consistency measure (incIN) with a threshold of 0.9.
Observation of the RoN measure — Relevance of Necessity — which evaluates how relevant a necessary condition is; that is, whether, besides being technically necessary (with high consistency), it also contributes substantially to explaining the result. The measure ranges from 0 to 1 (like consistency) and the closer to 1, the more relevant the condition is to explain the result.
Analysis of sufficient conditions. A condition is considered sufficient when X (condition) is present and Y (outcome) is also present (X⇒Y). That is, observing X and Y in the same case allows us to interpret that X is sufficient for Y, or that X implied Y in that case.
Construction of the truth table to obtain results from the analysis of sufficient configurations. It consists of grouping cases according to membership in the outcome of interest, considering all possible combinations among the causal conditions listed in the model, and identifying logical remainders (configurations for which there are no cases). Identify the output as a binary measure indicating whether a configuration is sufficient for the outcome of interest — in this study, satisfactory government actions.
Logical (Boolean) minimization. This process seeks to find, by contrasting the different configurations that led to the outcome, the simplest solutions free of redundancies, known as prime implicants. The process is repeated until all prime implicants are resolved.

Visualize the Data

Explore how these indices were used in the comparative analysis of countries

View Dashboard