Staking a claim

- What do we make of the rise in claims? "Last week’s rise in initial jobless claims to the psychologically important level of 500,000 has been met with heightened market scrutiny and suspicion. Initial claims, like most pieces of economic data, are subject to seasonal adjustment.
And, it is widely accepted that initial claims are notoriously difficult to adjust, particularly during the summer given school layoffs, holidays, and factory shutdowns in the auto sector. Now, a series of additional theories have surfaced attempting to explain the recent jump in initial claims:
• The extension of benefits back in July could be prompting Americans whose assistance ran out to file new claims in the regular program rather than file for the extended claim.
• A surge of Census laid-off workers filing for initial claims. Most of the Census layoffs have come during the summer.
• With the war in Iraq starting to wind down, we could be witnessing a rise in initial claim filings from our servicemen and women as they return home."
- If there is a bias in claims, it is upward and chronic "At the outset, let us say if there is a bias in the claims data, to quote the Department of Labor (DOL) official we spoke to, “it is an upward one” that is chronic. This is because an individual does not have to be eligible in order to file an initial claim. Initial claims are administered by states, which are required to take the claims form regardless of eligibility. The question that needs to be addressed is the extent to which the factors listed above are creating an abnormally large upward bias."
- There is a modest additional upward bias in initial claims "Our view is that these additional factors are having a limited upward impact on the initial claims data. The data from the Labor Department on actual payments allows us to create an implied approval rate, which supports our claim. This is important because while anyone can file and show up as an initial claim, what matters is if the government actually pays the claimant."
- Quantifying the upward bias "As of July, approvals ran at 69%. There is seasonality to this data and this is slightly below the 74% July average during the 2005-2007 period when the DOL was not making emergency benefits payments (Chart 1). So, taken literally this implies a 5% upward bias to the latest reading, or about 25,000. That would still mean claims are north of 450,000 and consistent with almost no growth in private employment."

Merrill Lynch Economic Commentary 20100826

No comments:

Post a Comment