Do State Indicators Trump National Ones for Predicting Economic Activity in the States? The Case of New York

October 2016

Since 2010, the Federal Reserve Bank of Philadelphia has published a leading index of economic activity for each state in the U.S. These leading indexes project the six-month growth rate of our state coincident indexes and have drawn the interest of the private sector, state governments, and academics.

Recently, at the request of the New York Department of Labor, we took a closer look at our leading index for New York to weigh the relative contributions of the state and national economic variables behind it.¹

Specifically, we conducted an empirical exercise to isolate the predictive performance of each variable over time. Knowing the importance of the state versus the national variables in our indexes is of potential interest for understanding the results of our monthly updates, for serving as a guide for ways to improve our statistical model, and for understanding the differences among states.

The key findings of our empirical analysis are:

New York state variables are more important than national variables in explaining variation in New York’s coincident index, suggesting that state variables are more important than national ones for our leading index.
New York state variables have been more important in the recent data than they were historically.

We thank the state of New York Department of Labor for inquiring about the relative weights of the variables in our model, prompting this formal analysis and, we intend, further work in this area.

View the Full Brief