Analytical framework

The simulator is based on a conditional labour demand function, in which employment is jointly determined by the level of output and the cost of labour. This framework allows for the quantification of how changes in productive activity translate into adjustments in labour demand, holding wages constant. For each occupation, an elasticity coefficient (β) is estimated, capturing the percentage change in employment associated with a one percent change in sectoral output.

Sectors and data sources

The tool covers two sectors directly associated with the hydrogen economy, selected based on data availability and comparability:

• Petroleum refining (CNAE 1921700)
• Electric power generation (CNAE 3511501)

Sectors are identified using the National Classification of Economic Activities (CNAE). Employment and wage data are drawn from the Annual Social Information Report (RAIS), which covers the entirety of Brazil's formal labour market. RAIS data are converted to monthly frequency to ensure compatibility with production indicators. Two variables are constructed from these data:

• Employment (L): number of active formal employment relationships, aggregated by sector, occupation, and month;
• Wages (W): average real wage, deflated using a common price index.

Output (Y) is measured as follows:

• Petroleum refining: physical production index from the Monthly Industrial Survey (PIM-PF, IBGE);
• Electric power generation: monthly electricity consumption data from the Energy Research Office (EPE), transformed into an index to ensure comparability with PIM-PF.

Occupations are classified at the two-digit level of the Brazilian Classification of Occupations (CBO-2). Within each sector, the occupations selected are those most frequently observed and most directly linked to the production process.

Reference period

The estimation covers January 2016 to December 2019. This interval was selected to capture employment and output dynamics in a relatively stable macroeconomic environment, avoiding the structural disruptions associated with the COVID-19 pandemic.

Limitations

Users should interpret results with the following constraints in mind:

• Static elasticities: The coefficients are estimated for the 2016–2019 period and reflect historical labour market conditions. They may not fully capture adjustment dynamics in structural transition scenarios, such as those associated with the scaling of a hydrogen economy;
• No substitution effects: The model does not account for potential substitution between labour and capital, or between occupations, that may result from technological change or shifts in production methods.