The best applied research is motivated by a timely problem. This point is echoed through the aims and scope of the many academic journals in applied economics. As a new challenge or perspective emerges, quality evidence-based analysis can provide valuable insight to the extent which it might influence the economy as we know it.
Unforeseen, emergent issues take a great deal of care to accurately reflect the subtle intricacies in stylized models with inevitably incomplete data. This takes significant time and effort.
However, in the past decade or so, large-scale and scoped structural models designed to tackle multi-dimensional issues such as climate change, energy, and food security have flourished. For instance, computable general equilibrium (CGE) models have many sectors and regions and characterize the interactions between agents in the entire global economy. These models are often the workhorses of integrated assessment models (IAMs) often used for climate change research and have been expanded to analyze issues such as energy, water, land-use, etc.
While these models may have been originally developed with a specific motivations, their wide-ranging capabilities lend themselves to an inherent “blue sky” perspective toward secondary motivations. This means that when a timely research problem emerges, the researchers do not need to engage in costly data collection model development because the existing model is quickly adapted. These can provide the timeliest responses to emerging research questions.
One of the more relevant examples of this is the shale oil and gas boom in the United States. Several CGE/IAM models needed only to exogenously expand oil and gas endowments to begin to explore the consequences of the unforeseen growth in oil and gas production in the US. Therefore, they can be first-movers in economic modeling of issues such as this.
The first-mover advantage refers to the gains accrued to the first entrant in a market by capturing a significant market share. The advantages might arise from proprietary information, preemption of scarce assets, and preventative switching costs; these are also relevant for first-movers in economic modeling.
In regard to proprietary information, one pervasive criticism of large-scale models is that they are often not completely documented - especially considering an academic article is roughly 7,000 words, and the documentation is often just given a citation. Also, there is little attention on validation (at least publicly documented) for many large-scale models. These points are akin to proprietary information where subsequent modelers cannot adequately replicate or improve upon the existing model design without recreating their own version. Also, without validation exercises, who is to say which model is an improvement?
First-mover models also preempt the scarce space in academic journals. Once a study is published in a particular journal the editor and reviewers may be less inclined to include subsequent researchers attempting to explore a similar or identical issue; it is no longer timely or novel. This may be especially true when the post-processing tools (e.g. statistical analysis) of the first-mover are far superior to that of an upstart researcher despite the upstart researcher selling a superior research product. Ask yourself which is more important: an application-motivated theory or error bars on generic theory?
Corollary to this, subsequent researchers will be required to invest additional time to “catch up” to the first-movers in terms of the slight variants in research problems. It may be the case that the costs of switching from the first-mover product are too large despite the subsequent movers having a superior product - I will continue to cite the first and most prominent study.
Another interesting feature of this problem is a combined first-mover effect that creates a sort of monopoly on the research space. The effect follows this timeline: 1) an interesting research problem emerges, 2) large-scale economic modelers are first-movers by slightly altering their existing models, and 3) the first-movers form an inter-model comparison consortium (e.g. AgMIP, EMF) to jointly publish on the research problem increasing the total market share. Inter-model comparisons sometimes are allotted entire special issues of academic journals. It is important to note that this is not an inherently good or bad thing, but simply that this effect exists.
For example, looking back at the shale oil and gas boom, several first-moving researchers have published work without simultaneously re-investigating the electricity sector. Many of these studies only consider the electricity sector as a single sector, so there is only implicit distinction between fossil fuel and renewable resources. Much of the shale gas will enter the electricity sector and will not be independent of the other policies that are technology-specific (e.g. renewable subsidies, nuclear phase-out). Neglecting a detailed treatment of electricity generation would likely give an incomplete view of how the shale oil and gas might affect the overall economy. Inter-model comparison studies compile the first-mover models to analyze their respective virtues regardless of their theoretical limitations.
The “blue sky” approach taken by many large-scale economic modelers is wonderfully flexible to make first-pass approximations of emerging issues in the timeliest way possible. Inter-model comparison studies are fantastic exercises to inform and advance the field by identifying best practices amongst the consortium. However, we must remember that there is always room for improvement from upstart researchers, we must be open to new ideas in subsequent studies even if they are not so timely any longer, and we must be vigilant to ensure the superior research product is the one that ultimately captures the market share.
- Jeffrey C. Peters