Uncovering Ghana’s Informal Sector: A Study of Household and Enterprise Surveys

A recent study led by the World Bank’s Development Data Group and Enterprise Analysis Unit, in collaboration with the Institute of Statistical, Social and Economic Research at the University of Ghana, offers critical insights into how informal enterprises are measured in low- and middle-income countries (LMICs). The research provides a comparative analysis of two major survey approaches, household surveys with an integrated module on enterprises (HS-IME) and the Informal Sector Enterprise Survey (ISES), to explore their effectiveness in capturing informal businesses' size, characteristics, and productivity. The study was conducted in the Ghanaian cities of Kumasi and Tamale, where both survey approaches were applied simultaneously, allowing for a direct comparison of methodologies in similar urban settings.

The Informal Economy: A Hidden Economic Force

In LMICs, informal enterprises form the economic backbone for large segments of the population, contributing an estimated 30% to 70% of GDP and employing up to 80% of the labor force. Despite their significance, these enterprises often go unregistered and unregulated, operating without business licenses, formal banking, or standardized bookkeeping. As a result, they are largely absent from national economic statistics. Accurately capturing the informal sector is therefore essential for effective policy planning and financial inclusion. This study aims to enhance understanding by examining the efficacy of HS-IME and ISES, two survey methodologies that approach informality from different angles.

Two Methods, One Objective

The HS-IME method samples households based on national census data and identifies enterprises through interviews with household members. It excels at capturing small-scale, mobile, or home-based businesses that might be invisible to the naked eye. In contrast, the ISES method uses adaptive cluster sampling (ACS), targeting geographic blocks and expanding to adjacent areas based on the density of informal enterprises observed. This area-based method allows for precise enumeration of visible enterprises, especially those located in commercial or densely populated areas, but may miss mobile or nocturnal operations that are not present during the day.

The divergence in methods produced striking differences in estimated enterprise counts. In Kumasi, HS-IME identified about 120,000 informal enterprises, while ISES estimated just 68,000. In Tamale, the gap was even wider, with HS-IME estimating 89,000 and ISES reporting only 21,000. These disparities stem from differences in sampling and observation: HS-IME is more inclusive of low-visibility businesses, while ISES tends to focus on spatially clustered enterprises during specific hours of the day.

What Makes Informal Enterprises Tick?

Despite their numerical differences, both surveys delivered remarkably consistent results when it came to the characteristics and performance correlates of informal businesses. Enterprises that owned business bank accounts, used mobile phones, operated outside of the home, and were active in the retail sector showed higher productivity in both datasets. These characteristics are widely associated with greater formality, access to capital, and customer engagement, all of which contribute to business growth and sustainability.

Descriptive analysis highlighted clear differences between the two cities. Enterprises in Kumasi appeared better resourced and more structured, with higher rates of bank account ownership and longer years of operation. Tamale’s enterprises, on the other hand, were generally smaller, more likely to be home-based, and more dependent on informal labor. Female employment was also more prominent in Tamale, though this was associated with lower productivity, a reflection of broader gender disparities in access to business resources.

Modeling Productivity in the Informal Sector

Using econometric analysis, the study further investigated the determinants of enterprise performance. Regressions on the log of sales and sales per worker, used as proxies for firm productivity, revealed strong consistency across the two survey approaches. Variables such as some workers, years in operation, use of phones, and bank account ownership emerged as key performance drivers. Interestingly, while a greater number of employees correlated positively with total sales, it showed a negative relationship with sales per worker, indicating that very small enterprises may operate more efficiently on a per capita basis.

Pooled regression analysis combining both datasets added depth to the findings. It showed that, even after controlling for location, ownership, and operational variables, enterprises surveyed via ISES had significantly lower productivity than those captured by HS-IME. This suggests that the ISES method may be better at reaching deeply informal and lower-performing firms. The difference also underscores how the choice of survey design can shape the statistical portrait of informality, potentially influencing how policies are targeted and resources allocated.

Toward Smarter Surveys and Better Policy

The study concludes with practical recommendations for improving survey designs and data accuracy. For ISES, adjustments to fieldwork timing and greater effort to account for mobile enterprises could mitigate undercounting. For HS-IME, incorporating questions to detect co-owned businesses across multiple households would reduce overestimation. Perhaps most importantly, the researchers advocate harmonizing question formats between the two surveys to enhance comparability. Taken together, the findings highlight the value of using both survey methods in tandem: HS-IME offers a broad view of informal enterprise activity, while ISES provides a closer look at specific business dynamics. When used together, they offer a more complete and nuanced picture of the informal sector, a vital step toward better policymaking in economies where informality is not the exception but the rule.