Printer Friendly

Forecasting likely job growth or decline with the net deficit technique.

This article examines the problem of forecasting changes in local job markets over time. The "net deficit" technique is suggested as a method for identifying periods of likely job growth or decline. The proposition introduced is that no local job market can sustain job growth above a long-term moving average. The longer that annual job growth is greater than the long-term moving average, the more likely it is that a period of job decline will occur. The type of moving average that might produce the best results is also defined. In addition to introducing the concept of the net deficit, the article, using the Houston area as an example, provides a detailed application of the technique.

Appraisers and others who attempt to make long-term planning, portfolio, and valuation decisions in real estate are constantly faced with the dilemma of ascertaining when job growth will occur in an area. Obviously, this type of forecasting is needed because job growth or decline is clearly linked to property absorption and changing property rents and values. When jobs are increasing in an area, rents and values are usually increasing, and vice versa. Cash flow forecasts should thus show an increase when jobs are expected to increase, and remain stable or show a decline if jobs are expected to decline.

Some analysts support their cash flow forecasts by obtaining job forecast information from local or state agencies such as chambers of commerce, state employment agencies, or university centers for business research. These forecasts are almost all based on mathematical statistics that contain inherent assumptions about such factors as interest rates and construction levels, and essentially extrapolate the past into the future while also usually containing subjective assumptions about the future. Most appraisers are aware that these forecasts increasingly have tended to be more wrong than right, and that they are changed on a regular basis. While there are many reasons for this, a primary one is that the U.S. economy does not influence the rapidly evolving global economy to the degree that it once did. Another reason is that the U.S. government's use of both fiscal and monetary policy appears to exert less influence than in the past. The old assumptions simply may not apply, and economic modeling for forecasting purposes thus may be less useful in the future. In addition, of course, the confidence of the public plays a major part in the equation. The bottom line, however, is that if an analyst is wrong, few remember the assumptions used while many remember that the final result is incorrect.

The purpose of this article is to present a new analytical tool that can be used to at least determine the most likely direction of changes in the job market and to ascertain at what approximate year those changes might occur. The "net deficit technique" recognizes that geographic areas experience cycles just as neighborhoods, companies, industries, and economies do. Geographic areas experience these cycles because an event or events bring jobs in great numbers to an area and then the construction cycle kicks in and accelerates the growth to levels that cannot be sustained over a long time. An event then causes jobs to leave the area, construction literally stops, and the area suffers a steep decline for some period of time. While the particular events can seldom be foretold in detail, the fact that they do occur is certain. During the down segment of such a cycle, local business leaders seek replacement jobs and at some point are successful, which cuases the cycle to repeat itself.

The net deficit technique is a tool that can be used in a straightforward fashion to help an analyst foretell the long-term job creation cycle, and more specifically to construct a net job deficit cycle. The technique is not complicated; an analyst simply has to know total employment in an area for a long period (at least 10 years), and have the ability to use any of the electronic spreadsheet programs on the market.


The net deficit technique was developed as a result of a study of the Houston economy that incorporated an attempt to relate job growth to property absorption. As is fairly well known, Houston experienced a boom-to-bust cycle during the 1980s (the oil job market was the event); similar events are occurring in other parts of the country. What is not generally known is that there was strong job growth in the area from 1976 to 1981, with 1981 being the peak year. There were six good years from 1976 to 1981 and six weak years from 1982 to 1987, followed by four good years from 1988 to 1991. The definition of a good year in the job market is one in which the job growth for that respective year was above the long-term average. This point relates to the central issues in the net deficit technique: Is job growth above the long-term average or below the long-term average? How is the long-term average computed? What is the cumulative difference between the long-term average and the net job growth for each year?

The Houston example, with an explanation of how each number was computed, can serve to demonstrate how the deficit technique can be used and how the technique can at least create a basis to test the forecasts of others.

The data and analysis presented in Table 1 contain information from the Houston Primary Metropolitan Statistical Area (PMSA) for the longest time period for which consistent data were available. Column A contains the average employment for the PMSA for each of the years listed. While there is monthly variation, the annual average is the most useful number because of the long-term planning horizon in the real estate business. Column B contains the net change in employment for each of the years listed. For example, the 64,900 for 1976 is the 1975 average employment of 993,000 subtracted from the 1976 average of 1,057,900.


Naturally the net change is positive when jobs increase and negative when jobs decline. Column C contains a long-term forward-moving average (LTFMA) of the net changes in column B, which begins with 1976. Because the LTFMA begins with 1976, the forward-moving average in 1976 is the change for that one year--the average of any one number is that number itself. The second number in column C is the average of the first two numbers in column B, the third in C is the average of the first three in B, and so on. Therefore the LTFMA is an average that starts from the beginning of the time period in question and moves forward in time, with each average being the average of all the previous net job changes for each of the respective years.

For example, the 1980 LTFMA of 81,200 is the average of the net changes in column B from 1976 to 1980. It is worth noting that the LTFMA has been declining in the Houston area. Column D contains a long-term backward-moving average (LTBMA), and repeats the process of column C in reverse. The 10,000 average for 1992 in column D is 10,000 because the average of the net change for 1992 of 10,000 is 10,000 (i.e., the average of 10,000 is 10,000). As the average begins with the most recent year and moves backward in time, each previous year's net change is added to the average and it is thus a moving average. For example, the LTBMA of 51,960 for 1988 is the average of the net job changes for the years 1988 to 1992. The 1992 average of 38,435 in column C is the same as the 1976 average in column D because each of these averages is an average of all the net changes in column B.

Because these LTFMAs and LTBMAs tend to be sloped as a function of whether the net changes have been increasing over the long term or, as is the case in Houston, decreasing, it is necessary to have some method of tempering this slope, or of smoothing the trend. To temper this slope, the averages of the LTFMAs and LTBMAs should be averaged for each of the respective years. These annual averages are contained in column E and this type of average is referred to as the long-term forward/backward- (LTFB) moving average. For example, the 1991 number of 30,281 in column E is the average of the averages for 1991 in columns C and D. (See Figure 1 for a graphical presentation of columns B, C, D, and E.) Figure 1 clearly reveals that a downward trend does exist as related to net job changes over time in the subject area.


The next step in the process is to subtract the LTFB moving average from the net job changes for each of the respective years under consideration. These calculations are presented in column F and these differences reveal on an annual basis whether the net job change for each year is above or below the LTFB moving average. If the net change in column B is greater than the LTFB moving average then a positive number results for that respective year in column F. For example, the 13,232 for 1976 indicates that the net job change in 1976 of 64,900 was 13,232 greater than the LTFB moving average of 51,668 for 1976. The data in column F reveal that for six years from 1976 to 1981 the area gained jobs faster than the long-term average for any extended period of time. This is because for such a gain to occur the rate of net job change would have to increase each year as the moving average continued to increase because there is a larger increase in net jobs gained with each successive year. An increasingly exponential job growth rate thus simply cannot continue in an area over a long period of time, just as a company cannot continue indefinitely to grow at an increasing rate.(1) Some event will occur to reduce the rate of net job change. In Houston it was the oil bust. In the Northeast it is the defense and electronics bust. While it is impossible to predict the exact nature of the event, some event will always occur just as Chaos Theory predicts.(2)

The data in column G represent a forward accumulation of the numbers in column F. The numbers in column G are referred to as the net job deficit and reveal that by 1981 there was an excess of almost 200,000 jobs above the moving average in the Houston area, which had accumulated over the boom. As expected, beginning in 1982 the net job changes for six years thus were below the LTFB moving average, and in some years the area even lost jobs. By 1985 the area had worked off the excess jobs and was near what might be referred to as "job equilibrium"; however, cycles seem always to overcorrect and this occurred in Houston. The main reason for the overcorrection was the almost complete stoppage of construction. The net job deficit turned negative in 1986, which meant that over time on a cumulative basis the area had created fewer jobs than the LTBF moving average, and had reached a negative peak in 1987 of almost 150,000 jobs. The area therefore was "short" 150,000 jobs when the total number of jobs was taken to be a function of the LTBF moving average. This was an indication that the stage was set for job growth above the moving average, and this is exactly what happened in the following years. From 1988 to 1991 the net changes in jobs each year were above the moving average and the deficit began to be worked off. In fact, the deficit would have been worked off by 1992 or 1993 if the national recession had not begun in 1990. (See Figure 2 for a graphical representation of columns B, E, and G.)


As indicated in Figure 2 and Table 1, column G, as the net deficit lessens there is an ever-increasing probability that some unknown event will occur to correct this excess of job creation over and above the LTBF moving average. Similarly, as the net deficit increases some event will spur growth above the long-term average. Of course in a large metropolitan area there would certainly be a number of events, even including relative cost factors such as high or low office rents caused by shortages or surpluses in the market, which may affect job growth as firms seek to lower the costs of operation. While the net deficit technique cannot predict the exact time for the changes, no model can predict with complete accuracy. This is why economists refer to their future numbers as forecasts. The net deficit technique, however, can indicate when job growth is probably going to increase or decrease over a given future three- to five-year period, and indicate as well how many jobs might be gained or lost before a zero job deficit will result. This information alone is often enough to make the difference between a winning and a losing situation.

The usefulness of this technique is to identify periods of excessive long-term job growth or decline and to recognize the probability that on a long-term basis, growth above the long-term average cannot continue and declines below the long-term average will not continue. No matter how optimistic the economic job growth forecasts are, above-average growth can continue for only so long, and analysts should compute the net job deficit numbers to check the reasonableness of the forecasts being used.


As we studied the Houston real estate market we found that the complete job cycle spanned over a decade.(3) Because some parts of the country are entering the downward phase of the cycle, it is critical to use this type of long-term trend analysis as appraisers working in weak areas attempt to estimate how low the market can really go and how long it might take to fully recover lost jobs and lost real estate values. The Houston experience has shown that even after all jobs have been regained, the values in some sectors of the real estate market will not recover 100%.

Those in the portfolio planning area should also find the net deficit technique useful. The net deficit technique can be used to help determine how many jobs might really be lost before a cycle turns positive, or at least to identify how many excess jobs have been created in the area. Because lost jobs mean fewer apartments rented, less office space rented, and less retail space rented, there may and probably will be a period of negative absorption and thus falling rents and values. Analysts in other parts of the United States might benefit from a study of the Houston and Texas experience to see how this cyclical trend occurred in one area.

The net deficit technique focuses on some measure of total jobs in the subject area; in Houston, the measure was total nonagricultural wage and salary employment. The reason for not attempting to examine each sector of the job market (e.g., construction, mining, retail trade) and instead to focus on total jobs, is that total jobs is the significant factor in regard to real estate property absorption. Obviously, each geographic area has different sectors of the job market and each of these different sectors has its individual cycle.

Further, local economies clearly are driven by what urban economists refer to as "basic" jobs, which are essentially jobs in local companies and businesses that export some portion of their product or service. This export is either to other parts of the United States or to a foreign country, and such jobs are called "exporters." To differentiate basic jobs from service jobs for purposes of the analysis presented here is no more necessary than to divide the cap rate into a return on and a return of money. It is recognized that sometimes basic jobs will grow faster than service jobs; this is when local restaurants and other retailers will make above-average profits. The market simply responds with more retailers, however, and a long-term balance between basic and service jobs in an area tends to occur. Further, sometimes service jobs grow faster than basic or exporter jobs. It would be of great benefit to know which businesses in an area are exporters, because these exporters (basic jobs) do drive the local economy; to know which of these businesses might grow or decline and thus cause the local economy to prosper or decline, however, is virtually impossible. The method presented here recognizes that some firms will fail and others take their places. Further, some event will cause these changes in a cycle and the event or events will be a function of Chaos Theory, which promulgates the idea that neither the event nor the timing of the event can be known.

It must also be recognized and remembered that the averages will change over time and that as the averages change, the indicated job deficit will change over time. The key is to remember that sustaining an above-average rate of growth for a long period is almost impossible, and that as the net deficit lessens the odds increase that a correction will occur.

While it is true that the ideas presented here have not been tested in numerous markets, the technique is based on common sense and logic and is easy to understand and use. In addition, the results of the study of the Houston market are remarkable in that the net job deficit cycle reflects the actual historical timing of the local cycle that occurred in the 1980s. Regardless of the lack of testing in dozens of markets, the technique does provide an additional tool for explaining changes in job markets, and can be useful.


Many market areas in the United States are experiencing reduced levels of job growth. As appraisers and others seek to use forecasts in the valuation process, there is a real need for information on how far the job market might fall and how long it might take to fully recover. Rent increases are almost impossible to sustain when vacancies exceed 10%, so as jobs decline vacancy rates increase and rents fall. In Houston, for example, it took six jobs to absorb an apartment unit in the 1980s and early 1990s, one job to absorb 173 square feet of office space, and one job to absorb 65 square feet of retail space. Each market has a ratio of jobs to property type use and by knowing these ratios, at the margin for the last 10 years or so, and how many jobs might be gained or lost, it is possible to estimate future vacancy rates. The net deficit technique is a tool to assist analysts in these estimates of job market changes. The technique is founded on the simple logic that job growth in an area will not, because it cannot, sustain either above- or below-average rates of change. At some point in the future any positive or negative excess of net job change, referred to here as the net job deficit, thus will be obviated. Just as rapidly expanding firms sooner or later face competition or even internal complacency, causing decline and eventual failure for some while others rebound, geographical areas go through cycles as well. Obviously, towns prosper when local exporters prosper and it is only logical that these exporters will sometimes falter and affect local towns. Eventually local people work to correct the situation and sooner or later the cycle turns up. The net job change cycle will always exist and the ideas presented here should help real estate analysts understand the cycle and thus be better able to foretell the direction of job growth in the future.

(1.)For example, according to one stock analyst, if Walmart continues to grow as it has during the past ten years or so, it will have virtually 100% of the retail market in the United States shortly after the year 2000. This obviously cannot happen.

(2.)According to Chaos Theory, some unknown future event will cause change or chaos in plans, cycles, and so on.

(3.)One of the writers has also studied the Lake Charles, Louisiana, job market and has found the same long-term time span for the complete job cycle.
COPYRIGHT 1993 The Appraisal Institute
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Smith, Charles A.; Forrest, William C.
Publication:Appraisal Journal
Date:Apr 1, 1993
Previous Article:Application of geographic information systems in site selection and location analysis.
Next Article:The role of a quality control checklist in avoiding appraisal report errors.

Related Articles
Consumer confidence, capital spending expected to stimulate economic growth.
Rating the consensus and Bureau 1992 forecasts.
Commercial real estate market demand growing.
2006 economic outlook: continued growth.
The UK's trade deficit reaches a high.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters