Gebretsadik Teshager Siltanu ^1* Carsten Smith-Hall ¹ Solomon Zena Walelign ^2,3

• ¹ University of Copenhagen, Copenhagen, Denmark
• ² International Centre for Evaluation and Development (ICED), Greater Accra, Ghana
• ³ School of Economics, College of Business and Economics, University of Gondar, Gondar, Amhara Region, Ethiopia

A recent body of literature has documented the importance of environmental income to rural households in the Global South. However, this literature has not been analyzed to establish which findings are robustwhat determines rural households' absolute and relative environmental income? We conducted a metaanalysis using published articles that measured environmental reliance from the Web of Science, Scopus, WorldCat.org, and MPDI databases. We examined the effect of socioeconomic, demographic, and resource site proximity variables on environmental income and reliance. We applied a meta-regression approach and included moderator variables such as sample size, survey frequency, and the types of journal to control for variations in effect estimates and assess risks of biases. We analyzed 112 studies published between 1996 and 2021 that together surveyed about 52,000 households in 35 countries. The findings confirmed that environmental income matters in total household income: environment, forest, and non-timber forest product reliance were, on average, 25±11%, 27±16%, and 27±16%. The level of reliance was moderated by region and the type of environmental products. On average, the proportions of significantly negative, positive, and statistically insignificant effect estimates were 25%, 18%, and 57%. All covariates, except distance to the resource sites, were weakly correlated with environmental income and reliance, indicating no globally robust covariates. Thus, policies and interventions should build on regional specificities.