Why can't economists agree?

There’s an old joke that says if ten economists are lined up and asked their forecasts for next year, there will be ten different answers. Indeed, I’ve been in programs with other economists presenting forecasts, and the joke isn’t far from reality.

Get ready for another round of “which economic forecast do you believe,” because with a new presidential term about to begin, analysis of economic proposals will hit the media fast and furious. And, a sure bet is, forecasts of the impacts of the proposals will be all over the board.
Some of these varying forecasts will be politically based. That’s to be expected. But there will also be serious analyses of policy proposals from well-trained and respected economists that don’t agree. This leads to a logical question: if the economists all received similar training, how can their forecasts be different?

One answer is that economists aren’t necessarily trained in the same way. Although there is a core of economic knowledge all students receive, at advanced levels of economics there are nuances of differences. For example, some programs put more emphasis on the ability of government programs to change economic outcomes. While other programs don’t discount the power of government in the economy, they put greater focus on the importance of private decision-makers.

In fact, the actions and reactions of private players in the economy can be an important reason why predictions of the impacts of government programs can vary among economists.

Here’s an example. Let’s say you read about a proposal for the federal government to increase taxes on high-income taxpayers by $1 billion. The revenues would then be used to fund job-training for workers whose jobs were displaced by the pandemic, with the ultimate goal of re-employing those workers.

 This is certainly a laudatory and important goal. But then let’s say you read about two conflicting studies of the proposal. One says it will be a great success, with the increase in jobs and incomes far exceeding the costs of the program. But the other study says the program’s success will be much less, with total net new jobs and incomes in the economy falling short of the program’s cost.
How could the results be so different? One reason could be the second study took account of “secondary effects” of the job training program, while the first didn’t. The secondary effects look at economic reactions and impacts from those who fund the program.
Here’s what I mean.  If high-income people are tapped to pay for the $1 billion job-training program, one view might be they won’t miss the money. Hence, there will be no reaction from those taxpayers.

That may be, but one reaction to consider is what these rich taxpayers would have done with the $1 billion if the government hadn’t taxed it away. It’s likely they would have done a combination of two things – spent some and invested some. Either way, spending and/or investing would have created jobs and incomes.   These jobs and incomes would therefore need to be subtracted from those estimated from the job-training program to derive a “net” gain in jobs and incomes.

Of course, the jobs and incomes created from the rich spending the $1 billion would not necessarily be the same jobs going to the unemployed workers participating in the job-training program. This would be an important factor to consider. Yet the significant point is, the “net” creation of jobs and incomes from the job-training program would certainly be less once the alternative spending of the $1 billion was considered.
Another reaction to consider is how the behavior of taxpayers now facing a higher tax rate might change, and how their reaction might impact the economy. Research shows that taxpayers are sensitive to tax rates.  Lower tax rates motivate people to earn more income – because they keep more of it – while higher tax rates cause people to reduce their pursuit of additional income – in this case because they keep less of it. As a result, if a government program is financed by higher tax rates, the potential loss of economic activity from those higher rates should be considered.

Good economic studies take this reaction into account. Such studies are termed “dynamic studies,” as compared to “static studies” which ignore the reaction. One problem is there is no settled, or widely-agreed upon, number that accounts for the reaction. One way to handle this issue is to conduct several analyses using different values. In the economics business, this is called “sensitivity analysis.”
The conclusion is, economic studies can reach different results depending on what the studies consider.  Studies which incorporate “reactions” of those footing the bill for programs can produce contrasting results from those not including the reactions.
So, as with most things, we have to peel back the cover on studies to see what is really going on with the calculations.  Only then can we decide what each study is telling us.

Walden is a William Neal Reynolds Distinguished Professor and Extension Economist in the Department of Agricultural and Resource Economics at North Carolina State University who teaches and writes on personal finance, economic outlook, and public policy