This paper is for regional workforce and economic development professionals, education providers, regional policy makers, and others with an interest in regional workforce and economic development. Our first task is to introduce the study of regional economics in a way that is clear and relatively free of jargon.
We will begin with an illustration. Think of a drip falling into a bucket of water. When it hits the water it creates a splash. Very quickly the splash turns into ripples that radiate out, bounce, rebound, and eventually settle back to equilibrium. Overall, the bucket will gain or lose water depending on the rate of water dripping into it and the rate of water escaping through leaks.
Regional economies are a lot like this, in that money flows in and out of them, and the new flow creates “ripples” of economic growth. Households, businesses, and governments are connected in a complex web of interdependent relationships based on producing, selling, purchasing, and taxing goods and services, and each activity in one place tends to have effects on other places—just like the ripples caused by a drip falling into the bucket.
Regional economists seek to understand and, as much as possible, explain these interactions and how a change (especially an increase or decrease in jobs and earnings) in one area might affect other areas.
Let’s take an example at the national (not regional) level:
- A sharp decline in auto sales will quickly lead to a decline in auto dealerships and auto manufacturing (e.g., less demand for cars = less production of cars).
- A decline in auto manufacturing will lead to a decline in business by those that supply goods and services to the manufacturers (e.g., parts suppliers).
- If sales continue to fall, some car dealerships, parts suppliers, and possibly even manufacturers will go out of business, which means a loss of jobs and money they generate from their sales, purchases, and salaries.
- With these jobs gone there will be less money circulating throughout the community and region.
- The ultimate effect will be less money being spent on all kinds of goods and services: groceries, doctors, houses, and countless other services, which means that local industries and services will eventually feel the effect of the downturn.
Each change in an economy produces additional ripples—sometimes large, sometimes small, that reverberate throughout the region and nation. If the ripples are small, many businesses will never actually notice them, or will be able to adjust to the fluctuations easily. However, if the ripples are large enough, many businesses could be capsized, which is where we find ourselves now.
Because we are talking about regions (not the entire nation) in this document, we’ll describe modeling initial events and their effects only inside one region at a time. In other words, we have to define where the walls of the bucket are. This is where the bucket analogy breaks down a bit, because the reality is so much more complex. We might instead imagine the flow of money among businesses as a network of water tanks and pipes, and regional boundaries are like walls that the pipes may or may not pass through. Some stay inside the regional “room” and others go straight out through the wall, but the vast majority divide into many small pipes, some of which feed back into tanks (which represent businesses, government, and consumers) within the room, while the rest go outside.
Regional economists have to decide where the walls are drawn, and then estimate and account for all the different flows in order to predict what the regional effect of a certain scenario will be—how much any new flow (or lack of flow) in one place will affect flows in other places, as well as the total amount of water (or wealth) in the room (region). A primary tool they use to do this is the regional input-output model.
A. What is an input-out (IO) model?
An input-output model is a way of representing the flow of money in an economy, primarily among industries, while also accounting for government, households, and regional imports and exports. An industry is a group of business establishments that share similar end-products (or services) and processes for creating those products/services. Once the flow is represented in the model, a user can introduce events that change the flow (such as loss or gain of jobs/sales in one industry) and simulate its effects on each industry in the region, as well as the region as a whole.
This complex web of transactions can be arranged according to a particular accounting system called “input-output accounts.” They are called “input-output” because a portion of the output (i.e., sales) of one industry will appear as the input (i.e., purchases) of other industries. These accounts track the flow of money from one entity to the next, and from them we can get a sense of the interconnectedness of the industries, households, and government entities that occupy a given geographic space and build models to automatically simulate and display these relationships. The input-output model therefore indicates how a change in one part of the economy will ultimately affect other parts based on these purchasing and selling relationships.
B. How are these models built?
The main source of all IO models in the United States is the Industry Economic Accounts—especially the Annual and Benchmark Input-Output Accounts—produced by the Bureau of Economic Analysis (BEA), which in turn depends on data from other federal agencies. These tables provide a summary of how industries produce and consume commodities at the national level, showing which industries produce and consume which commodities (including services), and how much. There are also some other rows and columns that account for money flows that are not strictly among industries, such as household consumption, government consumption, changes in inventories, exports and imports, compensation of employees, taxes, profits, and so on. In simple terms, these tables are a form of accounting—like a business’s balance sheet. They show how various sales and expenses add up.
These tables are then manipulated, combined, and customized for smaller regions of the country using each region’s own industry mix and other information. This process is called “regionalizing” the model, and it is crucial because we need to estimate how much of each industry’s inputs are obtained locally (within the region), and how much of each industry’s outputs are exported outside the region. Without a good estimate of this, we have no idea what kind of feedback loops are occurring. To use our previous illustration, we wouldn’t know how many pipes are directing water flow back to tanks inside the room and how many are going out through the wall. If a local widget maker purchases all or most of its inputs from outside the region, then the region will not benefit as much when that company grows. But if there is a local supplier network, the benefits (and the downsides, too, if the company were to fail) will be amplified locally.
The ideal regionalization method, of course, would be to survey every business in the region and ask about its suppliers and major clients, then classify and add up the responses by industry. But this is never feasible for many reasons, from cost to privacy concerns. So all the major models today depend heavily on non-survey techniques that use various regional data sources, including its industry mix, to estimate these values. Most models also allow some degree of additional customization if partial survey data are available.
The regionalization process results in a customized table for a region which shows what percentages of each industry’s “inputs” depend on the “outputs” from other industries. This table is the heart of any regional IO model.
C. Why is IO Important?
Any sort of regional planning, spending, or investment requires stakeholders to
- Express their goals clearly (i.e., create jobs, build new infrastructure, up-skill our workforce, attract new employers in industry A/B/C, etc.), and
- Assess the effectiveness of alternative actions in achieving those goals.
Furthermore, common goals for development projects include:
- increasing average earnings
- increasing economic diversity
- keeping more money in the local economy
- expanding various industry clusters
- increasing the overall quality of life
- developing new industry or occupational areas.
These needs are always present for long-term regional development. But they are particularly important in our current economic situation, with so many job losses and so much money being spent on projects and programs meant to spur economic growth and help people find new employment.
In each case an IO model should be an essential part of the region’s toolkit, allowing stakeholders to discover areas where the economy might benefit most efficiently from new investment, and conversely, areas where additional attention is likely to be of less importance.
With development goals in place, IO is used to identify the best approach. For example, suppose there are several groups competing for various infrastructure improvements within a state. The IO model is used to simulate the potential impact of each decision (in terms of job, area industry impacts, and increased revenues). With the model you can begin to determine which project would have the best impact on the region—based on increased regional earnings, job creation, and how well the investment complements the current economic base. As a result, alternative uses of existing resources or investments, and the impact of declining or growing industries can be assessed in a data-driven, objective way.
D. How are Multipliers Involved?
You’ll sometimes hear the term “multiplier” used in discussions of economic policy and modeling, usually in the context of job creation. Basically, a multiplier represents how much some aspect of a model will change in response to changes coming from “outside” the model. Going back to our “pipes” analogy, suppose we have a model of the flow among the water tanks and pipes in the room. The rate of flow from Tank A to Tank B in the room is a variable that’s entirely “inside” the model. But the model also depends on “outside” variables—in particular, how much total water is coming in. And if we change that amount permanently, we’ll change various rates of flow and water levels inside the room as well. We could then create multipliers to describe those effects in terms of the original change (final effect = the original change times the multiplier).
Economists use a lot of different multipliers to describe the effects of new private investment, government spending, and even tax cuts. In this document, we’re interested in industry-by-industry sales, jobs, and earnings multipliers in the context of regional economies. We’ll discuss more details in Section III.