The Occupational Supply Demand System (OSDS) offers two principal planning models as tools with which to guide training investments with indicators of balances or imbalances (i.e., skill shortages or surpluses) in occupational labor markets.

(A) The human resource accounting approach compares the projected total annual job openings (due to growth and replacement needs) for an occupational labor market with the recent output of program completers from related, structured training programs (of at least 300 class hours) at the sub-baccalaureate level for states, and at baccalaureate levels and above for the nation. It should be used primarily to identify occupational labor markets with skill surpluses, not shortages; because it utilizes only one source of labor supply information, that is, training completions data. At the national level, the leading example of applications of the human resource accounting model is the publication by the U.S. Bureau of Labor Statistics (BLS), entitled Occupational Projections and Training Data, 2008-2009 Edition, at www.bls.gov/emp/optd. The OSDS extends to the states some of the basic concepts of the Occupational Projections and Training Data publication of BLS, which was the statistical and research supplement to the Occupational Outlook Handbook (OOH), 2008-2009 Edition.

 With regard to the geographic units of analysis under the human resource accounting model, the higher geographic mobility of baccalaureate and above graduates limits comparisons of occupational employment projections and training data for BA/BS completers and above to the geographic unit of analysis of the nation. (With respect to empirical data about geographic mobility rates by educational level, please see The Performance Report For Ohio's Colleges And Universities, 2006, Ohio Board of Regents, at www.regents.ohio.gov/perfrpt/index.php). Because of the lower rates of geographic mobility of Associate Degree and below structured training program completers, statewide comparisons of total job openings and training program completers can provide useful labor market insights for sub-baccalaureate units of analysis (i.e., occupational labor markets).

 With this planning model, the training program investor relies upon the "Training, Other Qualifications, and Advancement" section of the occupational profiles in the Occupational Outlook Handbook (OOH) to help the program planner determine how well the training program completion data can serve as proxies for supply information. In the case of licensed, sub-baccalaureate occupations such as licensed practical nurses (LPN's), training completions data are good proxies for supply information; for the occupation of gardeners and groundskeepers, graduates from horticultural, structured training programs represent poor proxies for labor supply information, as explained and documented in the "Training, Other Qualifications, and Advancement" section of the OOH profiles for LPN's and gardeners and groundskeepers. gardeners and groundskeepers.

Because the training completions data represent only one source of labor supply, for many applications of the comparisons of occupational employment projections and training data, the human resource accounting model is indeterminate. For other applications, where the training data are good proxies of supply information, the human resource accounting model provides useful insights about the balances or imbalances in specific occupational labor markets. Most importantly, there have been instances with specific occupational labor markets, where the single labor supply source of structured training completers significantly exceeded the occupational demand estimates of total job openings, leading to a conclusion of a competitive job market - a conclusion which additional sources of labor supply resulting from unemployment, net occupational and geographic mobility, and new entrants into the labor market could only reinforce. In those instances, the training data become robust indicators of labor surpluses in an occupational labor market. The inconclusive comparisons of occupational projections of total job openings and training data are roughly analogous to the inconclusive areas for Durbin-Watson statistics about serial correlations, with supply/demand ratios that fall into determinate or indeterminate regions based on the particular characteristics of individual, occupational labor markets, the dynamics of which are described in the standardized, occupational profiles of the OOH.

(B) The model of occupational wage data over time analyzes wage data for occupations, and the industries in which occupations are heavily concentrated as critical labor inputs, over time from the Occupational Employment Statistics (OES) program, the National Compensation Surveys (NCS), and the Quarterly Census of Employment and Wages (QCEW) for industries. In a competitive labor market, ceteris paribus, the trends in occupational wages, for example, represent a summary of the results of the actions and reactions of both the supply- and demand-side actors in an occupational labor market, encompassing all sources of supply and demand.

The OES wage data come from surveys of employers, stratified to represent all employment size classes of firms. The NCS occupational wage data come from employers via a method of sampling called "probability proportional to employment size." As a result, as described by BLS, "the larger an establishment's employment, the greater its chance of selection" for the National Compensation Surveys. (Please see the U.S. Bureau of Labor Statistics web site about the NCS survey methodology at www.bls.gov/ncs/methodology.htm.)  For further, detailed information about the differences between the OES and NCS occupational wage data, please see the "Frequently Asked Questions," question #4, at the BLS site, stats.bls.gov/oes/oes_ques.htm.

With the occupational wage data over time, rapidly rising wages may indicate skill shortages, where other institutional factors such as unions or professional associations and credentials or licensing requirements do not artificially restrict labor supply. Conversely, the absence of significant increases in occupational wages over time suggests the lack of skill shortages.

Since the OES wage data surveys were designed as cross-sectional surveys, rather than longitudinal surveys, the review of OES wage data over time requires a cautious approach. For large employment states, when using the model of occupational wage data over time to identify likely skill shortages in the labor market, the analyst looks for cases where both the national and state OES wage data indicate percent change increases in wages for the same occupation and time period that are significantly greater than the percent change increase in the Consumer Price Index (CPI) for the same period of time, and significantly greater than the percent change, average occupational wage increases for the total all occupations for the same time period. Further, the labor market analyst (LMA) seeks confirmation of these OES national and state indicators of upward pressure on occupational wages from the National Compensation Surveys (NCS) for national and sub-state areas, regarding the same occupation of analysis and time period. For small employment states, the LMA may place greater emphasis in the analysis on OES and NCS wage data trends at the national level, which have smaller relative standard errors (RSE) of the occupational wage estimates and greater precision.

In addition, for those occupations with employment heavily concentrated in only one or two industries, the Quarterly Census of Employment and Wages (QCEW) is a useful source of complementary information about increases in industry total wages and average annual industry wage increases (i.e., total annual industry wages by annual average industry employment) over time at the national, state, metropolitan, and county levels. (Please see the QCEW web pages at www.bls.gov/cew.)