A large share of economic activity in the entire developing world - perhaps up to 50% - is conducted by informal and unofficial firms. Knowledge about informality is based on firm surveys that either randomly sample or list firms above a given size, or based on special surveys that require enumerators to find and interview informal establishments. All these approaches rely on enumerators locating “fixed establishments” where the informal firm is present permanently or for long periods. However, many firms and entrepreneurs in LICs are mobile, operating out of temporary stalls or structures. The degree of informality varies greatly even within the set of unregistered firms, and existing datasets have therefore rendered our knowledge of business activity in developing countries somewhat incomplete. Relying on administrative data, however, results in a hugely selected sample since informal firms are by definition excluded. Data covering small, informal firms remains rare.
The researchers plan to generate rich descriptive statistics of all 32,432 business enterprises located in our study area, including firms ranging from makeshift tea vendors to large multi-branch enterprises. Using datasets collected with prior grants, they will determine what fraction of establishments would have been missed in existing surveys or listings of informal firms that the firm dynamics literature frequently uses. Using administrative records matched to the geo-coded data, the researchers are able to examine patterns of formalization at the firm, industry, and area levels. In addition, information on dates of registration and firm age allows them to determine whether firms grow and formalize or whether the decision to formalize or not is in fact made early and is relatively inelastic, which means that it is not responsive to changes in the status quo.
Less than 40% of surveyed firms are tax registered and less than 10% made any VAT payment in 2012. Understanding the correlates of the decisions to register and pay will help to suggest potential interventions to improve formalization. Assessing possibilities for creating a longitudinal dataset (i.e. a dataset that follows the same entity for many points in time) is also of critical importance from a policy perspective: we know little about why firms fail or even if what we observe as a failure could in fact be a shift in the form of business activity. Understanding these issues is critical for designing appropriate policies to support SME growth.
