Govt unveils plans to build new Bermuda specific economic forecasting model
Government has announced plans to build a Bermuda specific model for all future economic forecasting after having missed badly with its growth predictions for 2010.
The report, which was released last month, revealed that the Island's gross domestic product (GDP) fell by four to five percent last year compared to the Ministry of Finance's forecast of 0.5 to one percent growth.
The announcement, which was made by Premier and Finance Minister Paula Cox in the House of Assembly last week, pointed out that economic forecasting in a small, relatively open economy was always problematic due to the significant impact of aggregate demand and supply on output.
“The reality is that no econometric model could have forecast the severity of the great recession on the Bermuda economy especially on the financial services industry which contracted by an extraordinary 27.5 percent,” said Ms Cox.
Economist Craig Simmons said that the biggest problem for economic thinking in Bermuda was the time lag in the release of data, with GDP data put on a year or more later; consumer price data with a three-month delay; while the latest retail sales figures were for December and the most recent balance of payment data was for the second quarter of 2010.
“Economic models are built on assumptions,” he said. “Quantitative, as opposed to qualitative, models use mathematical equations to simulate the economy, which is a nice way of saying that an equation is capable of describing human behaviour. That is a stretch, but economists believe that they can.
“For example, if an economist believes that a $1 increase in income leads people to consume an extra 80 cents worth of goods and services, then the marginal propensity to consume is 0.8. Based on that assumption, if we believe that Bermudians will be earning 10 percent fewer dollars in 2011, then consumer spending will decline by eight percent and the economy as a whole will contract by some multiple of the fall in income. There is a caveat: our predictions are based on nothing else changing. The real world doesn't operate that way, but that doesn't stop economists from surmising.”
Mr Simmons said that during the 1990s local policy makers relied on a model developed by Brian Archer, a British consultant and professor, to estimate the contribution or impact of tourism and international business on the economy, but it was an impact model as opposed to being used for making forecasts. Above all, he said the key was to understand the model being employed.
“When shopping for a boat (a model), one should know what it is one wants the boat (model) to do before one buys, and have an idea of the accessories (data) one will need to make the model useful,” he said.
“And equally important, policy makers who will be interpreting the results of the model should have a basic understanding of the model's assumptions and processes, otherwise it will be a case of blind trust.”