Zuñiga, Pluvia, et al. (2016). Productivity, Technology and Innovation Survey Barbados: 2013-2014. IDB Open Data. https://doi.org/10.60966/jh0bl3hy

1. English

What is this dataset about?

This dataset merges firm-level data from the Enterprise Survey 2010 and PROTEqIN 2014 in Barbados, focusing on technology in Barbados and how firms adopt innovation.

Which periods and firms does it cover?

The data pertain to 2013–2014 and are based on survey rounds conducted in 2010 and 2014. Firms from a variety of sectors are included.

Who produced this dataset?

Produced by the IDB’s Caribbean group, financed by IDB. The dataset includes technical documentation and questionnaires.

What kinds of variables does it include?

• Measures of firm outputs, inputs, productivity, technology use, and innovation activity
• Information on business environment, competition, trade, and firm characteristics
• Survey metadata, weights, and instrument documentation

How can this dataset be used?

• To analyze how technology in Barbados is adopted across firms
• To compare technology use and innovation with other Caribbean economies
• To identify barriers firms face in adopting new production technologies

When you see regression coefficients or marginal effects related to innovation adoption or productivity outcomes, here’s how to interpret them:

• Positive coefficient on innovation adoption (dummy or intensity):
  Suppose a coefficient is positive and statistically significant. In that case, it suggests that firms that adopt new technology (or increase their innovation activity) tend to have higher productivity (e.g., output per worker) than firms that do not adopt new technology or increase their innovation activity. For example, in the Caribbean PROTEqIN data, more innovative firms tend to perform better in productivity.

• Magnitude (elasticities vs effects):
  Coefficients can be elasticities (percent change) or level effects. If the coefficient is, say, 0.10 in a log-level specification, that implies a 10% productivity increase associated with a one-unit increase in the innovation variable (holding others constant).

• Control variables matter:
  To interpret the coefficient credibly, the regression often controls for firm size, age, sector, ownership, country fixed effects, capital per worker, and other relevant variables. That helps isolate the “innovation effect” from confounding factors.

• Causality and timing caution:
  A positive coefficient does not always imply that innovation causes productivity gains — reverse causality, selection, or omitted variables may be at play. Ideally, the dataset or model uses lagged innovation measures or panel methods to strengthen causal inference.

• Effect heterogeneity:
  The impact of innovation may differ across firm types (small vs. large) or sectors (export vs. domestic). Some studies show that foreign firms or exporters are more likely to adopt innovation and reap greater returns.

• Statistical significance & robustness:
  Pay attention to standard errors and significance levels (e.g., * p < 0.10, ** p < 0.05, *** p < 0.01). A coefficient must be robust across specifications to be trustworthy. Many studies of Caribbean firms use robust standard errors or cluster-robust standard errors.

• Interpreting ‘non-linear’ or threshold effects:
  Some models test whether innovation returns diminish or increase with scale, or whether there is a threshold above which innovation yields strongly positive gains. Be cautious in interpreting coefficients in non-linear models (e.g., interaction terms).

• Residuals and unexplained productivity (TFP):
  When regression models include inputs (labor, capital) and innovation variables, the residual terms often capture Total Factor Productivity (TFP). A high TFP residual suggests firms do more with the same inputs—perhaps due to managerial practices, organizational efficiency, or unobserved know-how.

What are the limitations of this dataset?

• It is cross-sectional (though “pseudo-panel”) and not a strict panel
• Self-reporting and measurement error may affect tech/innovation variables
• Differences in questionnaire rounds can introduce comparability issues

What is the current state of the tech industry in Barbados?

The Barbados Enterprise Survey shows that by 2010–2011, most firms had adopted basic digital tools: 85% reported using email to communicate with clients and suppliers, and 85% had their own website. By the 2013–2014 PROTEqIN survey, innovation-related questions revealed that firms were introducing new or significantly improved products and processes, though R&D spending remained modest. This indicates a tech industry in transition, with strong ICT adoption but limited investment in deep innovation.

How advanced is technology in Barbados?

Survey data highlights widespread ICT adoption but uneven advancement. In 2010, firms reported high-speed internet connections and cell phone use in operations. However, only a minority engaged in R&D or filed patents. By 2014, PROTEqIN data shows firms citing obstacles such as access to finance and an inadequately educated workforce, suggesting technology use was advanced in connectivity but lagging in innovation capacity.

What are the government initiatives supporting digital transformation in Barbados?

The PROTEqIN survey asked firms about public support for innovation and training. Some firms reported receiving assistance programs, though coverage was limited. This indicates government initiatives were present but not yet widespread, focusing on quality certification, export promotion, and training.

What is the Internet speed and telecommunications infrastructure reliability in Barbados?

Enterprise Survey data show that firms experienced power outages and occasional telecommunications issues. However, most firms reported having high-speed internet and using email/websites. Reliability issues were noted as moderate obstacles, but ICT penetration was high.

How do the Barbados vs Jamaica tech ecosystems compare?

Both countries participated in LACES and PROTEqIN. Barbados firms showed high ICT adoption but lower levels of innovation spending. Jamaica’s larger sample revealed more firms engaged in exports and facing competition. The comparison suggests Barbados had strong connectivity, but Jamaica’s ecosystem was larger and more diverse.