Here’s a dilemma in analyzing Cuba’s economy: To understand the dynamics, you need good data. But data on Cuba is sparse and sometimes inaccurate, making it tough to pinpoint the scope of change.

That was a key conclusion from a panel on economic measurement issues, held Aug. 4 during the 22nd annual conference of the Association for the Study of the Cuban Economy.

Jorge Perez-Lopez identified measurement problems in studying Cuba’s labor market. He said Cuba’s government statistics suggest a robust labor market, typically associated with a fast-growing economy. But visitors describe something else: an island where many people lack full-time work and hustle.

For example, Cuba’s official statistics says the unemployment rate averaged 2.5% in 2010. That’s the lowest rate in Latin America, and roughly one-third the regional average that year,  according to data provided to the UN Economic Commission for Latin America and the Caribbean (CEPAL in Spanish). 

 But that 2.5% rate masks many obvious realities. For example, he said, by counting only those looking for work, it excludes those who want a job but don’t sign up at a municipal job placement office to avoid being assigned to work as a farm laborer — often far from their home and in tough conditions.

The low rate also fails to count hundreds of thousands of government employees who hardly work. President Raúl Castro and other officials have estimated state payrolls are bloated by at least 1 million people, meaning that real unemployment should be much higher than 2.5%, said Pérez-López.

Indeed, CEPAL estimates that Cuba’s official unemployment rate of 7.7% would actually have been 19% in 1991, when the Soviet Union collapsed, if not for the bloated state payroll, he said (see related chart).

Cuba’s government data also undercounts average wages, said Pérez-López. The wage data considers only earnings from the formal sector, not the underground market. And it measures only wages paid in Cuban pesos, not payments in Cuba’s dollar-like currency. Many employers now pay in the higher-value convertible peso to motivate workers. Still, it’s hard to measure how much averages wages are underestimated, he said, “because data on wage payments in convertible pesos is not available.”

Measurements of Cuba’s economic output also raise concerns. Cuban data shows about two-thirds of the country’s activity comes from services, including medical missions sent overseas. Cuba’s dual-currency system helps Havana overstate the value of those services, some paid in hard currency or with petroleum. If Cuba adopted a single currency, the value of those services and total economic output might plunge.

By one rough estimate, the country’s total output for 2008 would be closer to $38 billion than the $70 billion that Havana reported, said Gabriel di Bella, an Argentine working with the International Monetary Fund.

So, what’s an analyst studying Cuba to do? Attendees gave these tips: Seek out as much data as possible from varied sources. Try to put the data in a broader context. Look for qualitative information, not just numbers. And recognize the limitations of measurement. In studying Cuba, said consultant Maria Werlau from the audience, “it’s important to take into consideration that the data we’re using may be very flawed.”