The index is based on changes in employment, housing starts, industrial production and house prices. Each geographic area is judged to be in recession, at risk of recession, recovering from recession, or expanding.
To minimize the effects of one-month spikes, the information is smoothed, usually by applying a three-month moving average. The result should be an indicator that shows marked trends in an area's economic activity, without making too much of temporary differences.
"We're more interested in persistence of trends than we are in one-month jumps in trends," said economist Andrew Gledhill of Moody's Analytics, which also sells reports and forecasts on every metro area.
A conundrum during a recovery
A reader asked, how can some of the metro areas on our msnbc.com map be in "recovery," if the statistics shown for that area are negative?
The answer: Moody's status assessment (in recession, in recovery, etc.) compares six-month periods. But the stats shown on the map are one-year comparisons. (In both cases, moving averages.)
You see where this is headed. If the sharpest decline was more than 6 months ago, but less than 12 months ago, then both these things will be true at the same time in areas beginning a recovery: stats will have improved over 6 months earlier, with the metro area showing up in the recovery category, and also the year-over-year stats will still be negative.
Here's more background to help you understand the index.
Geographic areas included
The index includes data for all 50 states, the District of Columbia and 384 metro areas. Those metropolitan statistical areas account for 84 percent of the U.S. population, as of 2008 census data. Although our charts are shortened, often a metro area includes multiple cities and even parts of more than one state. For example, what we label as the Allentown, Pa., metro area is known to statisticians as the Allentown-Bethlehem-Easton metro area, including parts of Pennsylvania and New Jersey.
Timing of the data
The first presentation of information includes data through January 2009. The staff at Moody's Analytics has gone back to apply the formula to data for every month beginning in January 1994. The information will be updated each month at adversity.msnbc.com. There's a six-week lag, so data for May are published in mid-July, for example.
The four parts of the index were chosen because they offer information about major parts of the economy: jobs, investment, industrial output and wealth. Together, they can indicate the trend of economic activity in an area, and each one can be updated frequently.
- Employment: a measure of job growth. The U.S. Bureau of Labor Statistics releases monthly payroll employment figures from its Current Employment Statistics survey, which is then seasonally adjusted by Moody's Analytics. The information is smoothed with a three-month moving average.
- Housing starts: a measure of investment. The U.S. Census Bureau releases monthly data on single-family housing permits. Moody's Analytics applies several factors to that information to estimate housings starts, including: adjustment for the lag time between a permit and the beginning of construction, adjustment for the share of permits that never move to construction, and an estimate of the share of housing starts that are not issued permits. The information is seasonally adjusted and smoothed with a three-month moving average.
- Industrial production: a measure of output. The U.S. Bureau of Economic Analysis releases monthly information on industrial production by type of industry, which is combined with employment estimates from Moody's Analytics to estimate industrial production at the state and metro level. The national changes affect a metro area's industrial production in proportion to each industry's share of manufacturing jobs in that metro area. For example, if the national increase is greater in manufacturers of automobile parts, then metro areas that are more dependent on auto-part manufacturing may see an increase. The information is smoothed with a three-month moving average.
- House prices: ameasure of wealth. This information is available quarterly, not monthly, but it is included in the index because of the importance of housing in the current recession. Information from the Federal Housing Finance Agency is converted to a monthly frequency by Moody's Analytics by spreading the quarterly change out over three months. The result is then adjusted to add a six-month delay, because it is generally believed that households react to changes to house prices after a time lag. This represents the wealth effect: As home values increase, homeowners react by spending. So rising house prices lead to more economic activity, not immediately but after a lag of about six months.
Adding it up
With all the components now in hand, the month-to-month changes for each are calculated.
Although wild movements are likely to be muted by the three-month moving averages, economic data still can be volatile, particularly at the level of a metro area. The final step in the process attempts to minimize these effects by adjusting the weight given to each component. Components that commonly show sporadic movement are weighted less, while smoother measures are given a higher weight. The weighting is different for each metro area. This mitigates, though does not eliminate, some of the potential for the Adversity Index to send false signals about changes in the health of any geography's economy.
To make the results easier to understand, states and metro areas are labeled each month. For an area to be deemed in recession, the six-month moving average of the index is lower than it was six months earlier. To be deemed in expansion, the opposite is true. The categories "at risk" and "recovery" are transition stages: At risk indicates that the economy is slipping from expansion toward recession, while recovery indicates movement from recession toward expansion.
If you have questions or comments on the Adversity Index, please send an e-mail.
© 2013 msnbc.com Reprints