In the past, people make predictions by extrapolation method from several unreliable data sources; therefore we didn’t get an overview of the economic activities as when we use Big Data.1
In early August 2015, the American press has brought the case of Federal Reserve System (FED) to light. They accidentally announced economic forecast for the next 5 years on their official website. These forecasts suggested that FED doesn’t want an economic crisis to occur before 2020. They also revealed the problems FED was concerned about, including economic security and how the economists using evaluation methods for economic data. In theory, FED will rely on the forecast data to make decisions, but the lack of coherence between the figures reported by the FED and what actually happened are not surprising. Whether accurate or not, at least the forecasting attempts made by economists had to produce specific results; but again there were few forecasts for the level of the Great Depression in 2008 - 2009, even after it had taken place, it was still missing forecasts with specific numbers. The problem lies in the fact that the most important indicators to measure, analyze, evaluate the economy are based on the data that is incomplete and not updated continuously. For example, the forecasters calculate real GDP based only on the estimated data on the beginning of each month in the quarterly GDP statistics, while these indexes are adjusted several times after that. As a result the final forecast will be delayed, far from reality. In the third quarter of 2008, only less than 30% of the professional forecasters joined the Survey of Professional Forecasters made correct predictions about the decline of the US GDP in the remaining months. In fact, US GDP fell by more than 8% in IV/2008. Economists, policy makers and business leaders will need the economic data that is more closely to the reality as a basis for their forecasts. Fortunately, thanks to the Internet boom, recently they are a source of new information: the big data collected by search engines and other Internet companies. Now people can track data in real-time jobs, then immediately recognize what areas are of greatest recruitment. Here is an example of one of the powerful tools to assess the labor market correctly and practically. "Billion Prices" Project at the Massachusetts Institute of Technology calculated inflation by using data in real time for online purchase from hundreds of retailers across the globe. "Google Price Index” also provides similar. Google Trends provides the information with search trend analysis on the Internet. Researchers can also extract information from social networks. For example, the hashtag #NFPGuesses on Twiter aggregates weekly predictions about the growth of non-farm payroll of the US NFP. Zillow, the online real estate services, helps to collect information on home sales and mortgage. Companies like SpaceKnow are using satellite images to monitor production activities. Unlike the surveys that are currently being used to serve the economic forecast, Big Data reflects the real-time behavior of the various economic sectors, thus indicating the change in background economy. For example, job hunting data, and job information can be used to predict the job market for the next month. If used properly, big data is capable of creating a revolution in the economic forecast. In the past, people make predictions by extrapolation method from several unreliable data sources; therefore we didn’t get an overview of the economic activities as when we use Big Data. In the era of Big Data, the challenge lies in the careful screening and analyzing large amounts of information. It will no longer simply collect data but the data is processed with a structure-specific data analysis.