Social Security has released updated statistics showing that the number of disability claims filed in calendar year 2019 was essentially the same as in 2018. This is a big change. In every year between 2011 and 2018 the number of new claims declined significantly, totaling a 32% decline over these years. Could this be an inflection point? It may be years until we know for sure.
You have to be careful in reading Social Security's table. First, two numbers are presented on number of claims filed, Field Office Receipts and Initial DDS Receipts. Field Office Receipts include many claims that were quickly denied on technical grounds such as the claimant failing to meet the earnings requirement. Changes in agency practices as well as the spread of internet filing of claims affect the number of technical denials. I think the Field Office Receipts numbers should be ignored. The Initial DDS Receipts numbers are the important ones. These are the cases that are actually adjudicated on the merits. Second, pay close attention to the footnotes which contain this caution:
Because the application data are tabulated on a weekly basis, some months include 5 weeks of data while others include only 4 weeks. This weekly method of tabulation accounts for much of the month-to-month variation in the monthly application data. This method also occasionally causes quarterly data to have either 12 or 14 weeks of data instead of 13 weeks, annual data may include an extra week of data.
This means that there's no point in paying attention to the monthly or even quarterly numbers. Even the yearly numbers may be a little distorted by having one week more or less than the prior year. Thus, even though the Initial DDS Receipts were recorded as being up by 9,195 in 2019 over 2018, I think it would be more accurate to say that there was almost no change from 2018 to 2019 since we don't know how many weeks were used for either 2018 or 2019.
Here's Social Security's graph of the data, which does suggest to me an inflection point but maybe that's what I want to see. Maybe the 2019 data will turn out to have been nothing more than a blip: