When I talk to people about why social entrepreneurship is important, I’m often asked whether I think all entrepreneurship is social. “Entrepreneurship creates jobs, and creating jobs is good,” the argument goes.

While there are some fundamental differences between traditional entrepreneurship and social entrepreneurship, I won’t deny that job creation is a generally good aspect of any type of entrepreneurship. With the economy stagnating and the unemployment rate hovering around 9%, the country could definitely use some more jobs. More employment would mean that people would have more money to save, spend, and invest. Those are important benefits. But is all job creation necessarily a good thing? Is there a way to measure the social impact of creating a job so that we can understand all the relevant costs and benefits?

Developing a metric for the value of job creation would be an important first step to make social entrepreneurship and traditional entrepreneurship more comparable. We would then be able to understand which traditional enterprises create significant social impact (positive or negative) and which social enterprises have the biggest effects. A recent This American Life episode on the tricky business of measuring job creation is well-worth a listen for anyone who wants to take a closer look at what people mean when they sax “x jobs were created last month.” To begin assessing the value of job creation, let’s start out by looking at what we can measure, or at least estimate.

First, there is the actual figure of number of jobs created, as measured by the increase in employment. The social impact of that change in employment might be measured in terms of the extra income someone earns and whether the job has benefits like health insurance or tuition support.

Another important factor is how other people benefit from the multiplier effect of creating that job. For example, say a parent in a family of four gets a new job. With the family’s extra income, they might be able to pay for a child’s education or to buy a more energy efficient car. Even people who don’t support families produce benefits for others when they get a new job. If a recent college graduate becomes employed, she might spend that money at the local corner store or coffee shop, thereby supporting the livelihoods of the people that work there. There is a ripple effect from creating even one job.

But not all jobs are created equal. The creation of a job at one company does not necessarily have the same social impact as the creation of a job at another company. For example, supporters of the proposed Keystone XL pipeline, which will run from Alberta to the eastern part of Texas, say that if built, the pipeline will create 20,000 jobs directly and about 120,000 jobs indirectly. Detractors argue that the pipeline’s costs do not outweigh the benefits of these added jobs. Specifically, they mention this safety risks and environmental impact of the pipeline. If the critics of the pipeline are right, the construction of the pipeline might result in negative social impact and more jobs. With cases like this one, it quickly becomes difficult to measure the social impact of creating a new job.

So what are the alternatives? How can we emphasize what we really want to maximize (social impact), if we can’t agree about how to measure it? The short answer is that it’s probably impossible without standardized metrics that all employers keep track of. But that doesn’t mean we can’t distinguish different types of job creation and try to promote job creation that creates the most social impact. What if we separated job creation figures into two groups: one that met certain basic criteria for socially beneficial job, and one that didn’t? Of course, deciding on what these criteria might be is easier said than done, but there might be some potential indicators that most people could agree upon. A job that had health benefits, paid a living wage, and was not created by a company that severely damages the environment or the communities in which it operates might be considered a socially beneficial job. Jobs that failed on some of these criteria would not earn this label. A metric like this would be fairly rudimentary, but it might add some perspective as to which jobs are producing a greater social impact than others.

What this discussion about measuring the social impact of a job ignores, however, is that some of the most important aspects of employment can’t be measured. Russell Muirhead describes the importance of finding fitting work in his wonderful book Just Work, and argues that the dignity and satisfaction of working on a job that contributes to the common good are fundamental parts of good work. Although attempts to measure the social value of a job are important, we shouldn’t lose sight of these less tangible benefits. By simply looking at basic employment figures, we lose sight of how much jobs vary in terms of how much social good they create. We should prioritize jobs that promote the common good and produce internal goods that are harder to measure like self-worth and the opportunity to develop one’s capabilities as a human being. Until we take a closer look at the differences between jobs, the term “jobs created” will remain vague and won’t tell us much about what actually matters.