I’ve been running my own econometric model of the U.S. economy for almost 30 years now. The basic structure is simple. You start by forecasting the components of demand, that is to say, consumption, investment, trade and government spending. This gives you an initial projection of real GDP growth. You then feed this into labor market equations, along with some demographic assumptions, to forecast the growth in jobs, the unemployment rate and wage growth. All of this, along with assumptions about energy prices and the dollar, then drive forecasts of inflation. Given this outlook for growth and inflation, you make an assumption about the path for the federal funds rate and then run forecasts of other interest rates. With all of this in hand, you can forecast productivity, corporate profits, the federal budget deficit and household net worth. And then you go back to the start to see how all these changes impact your original demand forecast. You repeat the process until you arrive at a reasonably consistent solution.