Development and Validation of a Real-Time Happiness Index Using Google Trends™

Abstract It is well-established that a positive relationship exists between happiness and the economic outcomes of a country. Traditionally, surveys have been the main method for measuring happiness, but they face challenges such as “survey fatigue”, high costs, time delays, and the fluctuating nature of happiness. Addressing these challenges of survey data, Big Data from sources like Google Trends™ and social media is now being used to complement surveys and provide policymakers with more timely insights into well-being. In recent years, Google Trends™ data has been leveraged to discern trends in mental health, including anxiety and loneliness, and construct robust predictors of subjective well-being composite categories. We aim to construct the first comprehensive, near real-time measure of population-level happiness using information-seeking query data extracted continuously using Google Trends™. We use a basket of English-language emotion words suggested to capture positive and negative affect and apply machine learning algorithms—XGBoost and ElasticNet—to identify the most important words and their weight in estimating happiness. We demonstrate our methodology using data from the United Kingdom and test its cross-country applicability in the Netherlands by translating the emotion words into Dutch. Lastly, we improve the fit for the Netherlands by incorporating country-specific emotion words. Evaluating the accuracy of our estimated happiness in countries against survey data, we find a very good fit with very low error metrics. Adding country-specific words improves the fit statistics. Our suggested innovative methodology demonstrates that emotion words extracted from Google Trends™ can accurately estimate a country’s level of happiness.