A OCDE publicou uma interessante nota técnica acerca do tamanho da economia não observada nos números do PIB. O tema é bastante árido, mas tendo em conta a quantidade de comentários que surgiram depois de o INE anunciar que ia começar a produzir estimativas explícitas para o valor económico da prostituição e do tráfico de drogas, vale a pena deixar aqui o link: The non observed economy in the system of national accounts. Em baixo, a avaliação da OCDE da “economia ilegal” que frequentemente aparece nas notícias.
It follows from the identification and calibration problems described above, that these approaches do not allow delineating where their “shadow economy” starts and ends. Although the most commonly used definition is close to the underground economy in the Handbook – and this excludes both illegal and informal economies – the explanatory variables and the indicator variables are still very much likely to be correlated with these excluded phenomena, and perhaps also with activities in the observed economy. This again shows the somewhat arbitrary nature of the connections that are claimed to exist between the models and the NOE to be measured.
In addition to the limits of the macro-econometric approaches outlined above, there are several other recurrent issues in the NOE macro-econometric research regarding the use of the data, especially GDP, and/or the interpretation and presentation of results. Usually, the relevant research disregards the fact that published GDP data already contains estimates of the NOE. Consequently, their estimates are presented in percentages of the official GDP as the entirety of the “shadow economy”, implying that the official GDP figures do not capture any of the NOE. However, as explained in the main text of this note, official GDP data for most countries, including all OECD countries, already include adjustments for the NOE. Hence, there is an implicit double counting in the results derived from macro-econometric models which can lead to large overestimates of the true size of the NOE.