I observe economists engaged in bitter debates over whether economic growth is helping the typical American. One reason that these arguments strike me as futile is that I believe that the basic indicators are fundamentally flawed.
The most widely-cited indicator of economic performance is GDP per worker, along with related measures such as average wages adjusted to reflect changes in the cost of buying a particular market basket of goods and services. Much has been written about the merits and defects of GDP. See Diane Coyle, GDP: A Brief but Affectionate History.
In my view, as we continue to evolve toward a post-industrial economy, GDP is becoming less and less reliable as an indicator. In The Value of Nothing, I argued that value is socially constructed. From that perspective, the treatment of GDP as an objective measure is a false pretense.
What about happiness?
Some social scientists have proposed using subjective happiness as an indicator of economic performance. I agree with the intention, but not with their method of measurement.
The usual approach to measuring happiness is to take surveys. People are asked to rate their happiness on a scale, say 1 to 7. My concern with this approach is that although it gets you numbers, we do not really know what they mean.
Subjective happiness is not an absolute. To respond to a happiness survey, you necessarily assess your happiness relative to some baseline. But what is the baseline? Your memory of how you were feeling an hour ago? A year ago? Or is your baseline your beliefs about the happiness of other people, based on their income, status, and appearance?
When you fill out a happiness survey, are you supposed to describe how happy you have been today, over the last month, or over your entire life to date? Or are you supposed to describe how happy you are at this exact moment? Or how happy you are anticipating next week, when you are going in for a colonoscopy?
Even if happiness were well measured, the determinants of happiness include non-economic factors. If we are trying to evaluate economic policy and economic outcomes, happiness is too broad, and GDP is conceptually more appropriate.
Occupational satisfaction index
I propose that, instead of using GDP or overall subjective happiness, we should use occupational satisfaction as the broad indicator of economic performance. Occupational satisfaction is the core economic component of happiness. Unlike GDP, which is rooted in a materialistic understanding of value, occupational satisfaction reflects the understanding that value is subjective.
There are many drivers of job satisfaction. These include monetary compensation, status, work relationships, sense of accomplishment, sense of meaning, sense of control, sense of structure, fit with individual talents and interests, and more.
As with happiness, occupational satisfaction is comparative. A good friend of mine points out that many people become annoyed thinking about others who get more income and status out of less skills and effort than we contribute. But we rarely stop to consider that there are others who contribute more skills and effort but attain lower income and status. Perhaps if we spent more time remembering the people who are underpaid relative to ourselves, we would experience more job satisfaction.
I am not familiar with the literature on job satisfaction surveys. One approach to asking about job satisfaction would be to assess the person’s willingness to quit or to change jobs. How often do you think about quitting? Based on what you know of working conditions at another firm where you could imagine yourself working, how much would they have to pay over (or under) your current salary in order for you to be willing to move?
The challenge with assessing job satisfaction this way is that it may be difficult for the individual to visualize the comparable employer in a holistic manner. So I would be inclined toward a different approach.
I suggest listing all of the possible drivers of job satisfaction, including monetary compensation, quality of work relationships, travel requirements, and so forth. Have the individual weight the importance of each characteristic, and have the individual indicate his or her level of satisfaction with that characteristic in his current job.
For each characteristic, the employee should be instructed: assess satisfaction relative to what that you believe would prevail at another organization where you could picture yourself working. Asking the question this way would serve to clarify the benchmark for the respondent. I might note that if all of the relevant organizations were to improve (or deteriorate) at the same rate in all respects, this relative standard of comparison would show everyone’s job satisfaction standing still. My hope is that in practice this bias toward showing stability would be small and that researchers would find ways to adjust for it.
The weights on importance should be made to add up to 100 percent. The weighted average of the satisfaction rankings for each component then becomes the individual’s job satisfaction index.
My hypothesis is that presenting the survey respondent with a list of characteristics would be more effective than simply saying, “rate your overall job satisfaction on a scale of 1 to 7.” I think that breaking the question down into characteristics would encourage an answer that incorporates more thought. But that hypothesis could be wrong.
It would take a lot of trial and error to develop an effective job satisfaction survey of this sort. The process would include testing out various characteristics and different wording to describe each characteristic. It would involve getting feedback from those who fill out the survey about ways to improve it. Researchers probably would discover some issues that they had not considered at first.
Once the survey instrument has been established, then it can be widely administered to a large sample on an annual basis. Taking the average of each individual’s job satisfaction index (including the values for those without paid work) would yield a national occupational satisfaction index.
Gains in health and safety
If we had a national occupational satisfaction index, then I speculate that it would show a positive trend in two respects and a negative trend in another. Jobs are getting safer and the chances of obtaining a good fit for the individual are higher. But the decline in labor force participation suggests that for some people it has become more difficult to find satisfying paid work.
The data cited earlier show dramatic improvement in occupational health and safety. The rate of reported workplace injuries and illnesses fell from 8.4 per 100 workers in 1994 to 2.9 per 100 in 2016. This represents a drop of more than 2/3 over this brief period. Although this may reflect better workplace safety within occupations, it probably is due more to trends in the mix of occupations, with a shift away from physical labor toward lower-risk jobs.
While all of us enjoy physical activity up to a point, we are better off when we can set our own limits on our exertions. If you tend to your garden as a hobby, then you can stop if your back starts to hurt. Workers who pick crops for a living do not have that choice.
Physical labor, of the sort carried out by farmhands, construction workers, miners, loggers, and factory workers, tends to expose the body to regular punishment and sudden risk. From a health and safety perspective, the decline of those jobs is a boon. Assuming that workers value health and safety highly, this means that the occupational satisfaction index would have gone up.
Another aspect of job satisfaction that I suspect has gone up is the ability to find work that offers meaningful use of one’s skills. One sign that this has increased is that more people are working at non-profits, which means that more people are choosing to work based on mission. (Note that I believe that for-profit firms have missions that are often at least as compelling and that they tend to serve their missions more effectively; but the point that individuals have choices still stands.) Another reason to suspect that skill-matching has become more refined is that there has been a proliferation of new types of jobs, from specialty yoga instructors to web designers.
The decline in labor force participation shows that not every trend in job satisfaction is positive. If people thought that the available jobs were sufficiently rewarding, then fewer people would have stopped looking for work.
Economic heresy
In some ways, the concept of an occupational satisfaction index as an indicator of economic performance represents heresy. In conventional economics work is a bad, not a good. Also, economists take “satisfaction” for granted: when an individual makes a choice, then by definition the individual is satisfied with that choice.
From the perspective that work is a bad, the value of a job is measured by monetary compensation. But for measuring job satisfaction, even using monetary compensation is problematic, because non-wage benefits, such as health insurance, may be valued by workers at either much more or much less than what they cost employers.
Moreover, much of what would account for job satisfaction would go unmeasured, because these drivers are not included in monetary compensation. Safety on the job is unmeasured compensation. Congenial co-workers are unmeasured compensation. A meaningful organizational mission is unmeasured compensation.
In standard economics, a choice is a choice, and making a choice means that you are (just) satisfied. If your choice just barely makes you satisfied, then price is a good measure of value. But what if you are more than just barely satisfied?
Economists do have a theoretical concept, known as consumer’s surplus, defined as the satisfaction that you get that is above and beyond what you would need to just barely be willing to choose to purchase a good at its selling price. Similarly, receiving compensation from a firm that is more than what you would require as a minimum to work is known as producer’s surplus. But this surplus is not something that is measured in GDP or related indicators.
Standard economics is focused on what seems tangible. It assumes that people value goods, services, and leisure. GDP attempts to put a value on goods and services, but this is a futile attempt to arrive at a tangible measure for an increasingly intangible economy.
From the standard perspective, worker satisfaction is measured as monetary compensation relative to the cost of some standardized set of consumer goods and services. This measure is treated as objective and definitive.
In fact, although monetary compensation is important to most people, it is not the only aspect of a job that they value. Taking into account all of the non-monetary aspects, job satisfaction and monetary compensation may be only weakly correlated.
In summary an occupational satisfaction index would have the following advantages over current economic indicators:
— It would take into account intangible drivers of satisfaction on the job, which are increasingly important.
— It would take into account producers’ surplus for workers, which may be very high in some cases.
While an occupational satisfaction index would not be a perfect indicator of economic trends, it could be better than what we have now.