Indian national elections are here. India chooses members of parliament – Lok Sabha. And then the parliamentarians choose the Prime Minister. The process is similar to other parliamentary democracies. The Indian elections will be conducted in the months of April and May, and the electoral outcome will be known about May 23rd. There are 543 parliamentary seats. A government must enjoy the majority support in the parliament. There are many polls predicting the outcomes of elections in India. But I have discomfort in accepting polling results in India. For most part, it appears that they get the numbers wrong. The error-prone forecasting of Indian elections extends to forecasting in other domains in India, including business and economic forecasting. In general, forecasting is difficult and requires careful attention to data collection, surveys, modeling and statistical analyses. Probabilistic/Stochastic modeling is critical. This Scholarly article¹ presents many of the challenges in forecasting. These challenges are even more amplified in the Indian context. Here, as an illustration, we examine political forecasting. The same polling companies and pollsters – for most part – also provide business and economic forecasting in India.Read Full Article