This paper explores macro-, and meso-economic forces shaping search and explorativebehaviour at the firm level. In particular, we focus on top-down selection mechanisms (bothdirect and indirect) that modify the rate and direction of technological change, as well asthe nature of technological learning at the micro-economic level.Studies on economic development in developing countries focus increasingly on theconcepts of technological learning and technology systems. It is well accepted thatfundamental causes of differences in long term economic performance expressestechnological asymmetries at the firm and sectoral level (Dosi, Pavitt and Soete, 1990;Fagerberg, 1994). Given that technological learning is embedded in production chains,knowledge networks, and institutions, patterns of accumulation of technologicalcapabilities depend in turn on related systems of innovation (Lundvall, 1992; Nelson, 1993).Opportunities for catching-up and narrowing technological asymmetries depend as well onthe particular innovation regime, the rules govern search and innovative behaviour that in aparticular industry and in a particular time (Nelson and Winter, 1982; Malerba andOrsenigo, 1995).However, macro-to-micro causality has been largely excluded in these fundamentalexplanations of economic performance (see Cimoli and Katz, 2002). Macroeconomicenvironments determining levels and growth rates of key variables can influence directlythe rates of technology absorption and local innovation through traditional channels likerelative prices of capital, capital turnover rates, market size and growth rates, etc. Next tothese quantitative impacts, we argue, macroeconomic settings can also affect the nature oftechnical change, by influencing market- and firm-level selection mechanisms that governan economy's capacity of generating variety.Following Nelson and Winter (1982) model of search and selection, we develop an analyticframework that links explicitly micro-level behaviour and learning, with macroeconomicselection mechanisms. In turn, we show how macro-to-micro causation can biastechnological behaviour into certain directions, reinforcing cumulative causation and path-dependent features of technological learning. Next, we examine the conditions under whichadverse selection environments can generate development traps of slow economic growth,and slow variety generation. In circumstances where active competition and markettransference mechanisms operate in a relatively fluid manner across borders, economicresilience is crucial for preventing economic systems to rely purely on static advantagesthat may result socially and environmentally deleterious in the long term. Finally, we arguehow the learning restrictions set by astringent selection environments can and have beenovercome under certain innovation regimes.