I know–why predict 2015 outcomes. That’s last year, right?
Right. But the Census Bureau publishes estimates of household income and poverty results with a lag. This year, their results for 2015 will come out on Sept. 13. I’ll write them up that day, as is my wont, but for reasons I can’t explain, I always like to try to forecast them.
My track record isn’t great. Last year, I was fairly confident that poverty fell in 2014, but it held at 14.8%, meaning that’s the share of the population with 2014 incomes below the poverty thresholds (plural because the thresholds change by family size). However, the 2014 results were somewhat ambiguous because Census did some redesigning of their income questions. There may have been some comparability issues with the samples used to compute the changes in poverty and income.
(I took some solace in that my forecasts matched the results from a different Census survey–the American Community Survey–which comes out a couple of days after the one discussed here. I think the ACS data were more reliable last year.)
This year, I’m moderately confident that I’m right, at least directionally. My forecasts are:
–poverty fell half-a-percentage point, from 14.8% to 14.3%.
–real median HH income rose 1%.
The poverty change would be statistically significant; the income change would not be.
For a variety of reasons, we could see more of an upside than a downside surprise–i.e., a greater fall in poverty and rise in income than I’m predicting. For one, I think the survey changes may have missed the decline in 2014, so there’s some pent-up juice there. For another, inflation was almost 0 last year, and that turns out to matter in these data.
The poverty thresholds are adjusted by the rate of inflation. So if inflation is very low, the thresholds hardly change at all. Meanwhile, people’s nominal incomes tend to rise year-over-year, especially when the job market’s tightening, as it was in 2015. That should push up the real HH median income and bring down the poverty rate.
My model runs off of mostly inflation and labor market variables, and since I have these outcomes for 2015, I can make the prediction.
So, based on very low inflation and the growth in jobs and real wages last year, I think the results will be a marked improvement over last year’s. That doesn’t imply happy days, ftr. After six years of recovery, both variables will still have not returned to their pre-recession levels back in 2007.
The politics of this are always interesting in an election year. The R’s just finished a convention that painted what I’d bet is the most negative picture of America ever to come out of party’s nomination process. If I’m in the ballpark, the results may cause some dissonance, though that assumes an acceptance of data/facts–a strong assumption, indeed.