INTRODUCTION:

IMPLAN economists are often asked about best practices when modeling changes in utility prices and rebates on energy purchases.  This article gives some general guidance on how to best proceed when investigating these situations.  

DETAILS:

The first step in looking at changes in energy prices is to determine who is the producer and the consumer. In IMPLAN, utilities are produced by both Industries (Sectors 41-51) and Government (Sectors 519, 522, & 525).  Consumers of utilities can be Households, Industries, and Government.

Perhaps you want to model a price change in the Commodity, not knowing if it will be produced by a firm or the government.  If this is the case, consider using a Commodity such as 3041 – Electricity.

Next, you will need to make assumptions about the price increases or decreases that will be modeled – who will be affected and if that change is positive or negative.  This will frame your analysis.

  HOUSEHOLDS:

A popular case to examine is modeling increased income that households would see from a decrease in utility pricing or from utility rebates.  To model this, a Household Income Change (HHIC) is a great option. The HHIC will remove personal taxes and savings before calculating the results.  

Any changes to household spending will be at the same rate across all spending categories.  This means that even though the savings might be spent on increased retail shopping or entertainment, the spending pattern will allocate the savings across all household purchases, which doesn’t always make sense.  Just because someone saved money on their utility bill doesn’t mean their rent increased. Each analyst must determine how the savings will be spent.

In the case of a rate hike, you would see a reduction or negative impact in households.  Be careful in this case as well. Giving a household less disposable income doesn’t necessarily mean they won’t pay their student loans or buy groceries. They may instead cut their utility usage or decrease their entertainment budget. If the increase is small enough it may have little to no impact on household spending for any group. 

Remember, different households may respond differently, depending on the size of the rate increase, their needs, and income level. A larger increase will likely affect the spending of lower income households but may not affect the spending of higher income households as these payments may come from savings (leakages in IMPLAN). 

 INDUSTRIES:

PRODUCING

As a producer, Industries can be positively or negatively affected by price changes.  If the cost of inputs skyrockets or bottoms out, these price changes can’t immediately be passed on to the consumer, so short-term losses may occur.

Changes in Industries can be modeled through:

• Sector 41 – Electric power generation – Hydroelectric
• Sector 42 – Electric power generation – Fossil fuel
• Sector 43 – Electric power generation – Nuclear
• Sector 44 – Electric power generation – Solar
• Sector 45 – Electric power generation – Wind
• Sector 46 – Electric power generation – Geothermal
• Sector 47 – Electric power generation – Biomass
• Sector 48 – Electric power generation – All other
• Sector 49 – Electric power transmission and distribution
• Sector 50 – Natural gas distribution
• Sector 51 – Water, sewage and other systems

CONSUMING

Estimating a change in Industry price or profit is outside of the scope of Input-Output analysis. Therefore, the analyst must make assumptions about where the price changes would occur and how to model them.

For a price reduction or rebate, the likelihood is that this savings would go into Other Property Income (OPI).  OPI in IMPLAN is treated as leakage. However, you may know that this savings is allowing the affected Industries to hire more staff or produce more widgets. This can be modeled in IMPLAN.

If you want to assume that the businesses cannot control the price of their products, any increase in utility costs could show a loss in profit. It might even be enough to shutter some businesses altogether.  More likely, however, they will just raise the price of their products to offset the higher bills.

  GOVERNMENT:

PRODUCING

As a producer of utilities in many areas, the changing cost of inputs can affect the operating margin of government-run utilities.  Changes to government utility production can be modeled through one of the government utility Sectors:

• Sector 519 – Federal electric utilities
• Sector 522 – State government electric utilities
• Sector 525 – Local government electric utilities

CONSUMING

Governments can be affected negatively by increased utility costs, as well.  More money being spent on electricity, for example, decreases the overall budget and might lead to program cuts or elimination.  Often this is just in the short-run, however, as increased costs will likely be passed onto the consumer.

Governments can see savings as well when utility costs decrease. Analysts will need to determine how to model these cost savings. Knowing how the savings will be spent is key to being able to modeling the impact. If a new school is built with the savings, then a simple construction Event is appropriate.  Perhaps you may only know which government entity will spend the savings, in which case you should choose from the Institutional Spending Patterns.  When using an Institutional Spending Pattern, remember to modify it to remove the purchases of Commodities related to the utility generation you are examining.

  CONSTRUCTION:

Although temporary in nature, construction projects conducted by utility companies are often large and costly.  Consider modeling this as a separate impact from operations or price changes.

  NET ANALYSIS:

If a government entity is considering building a new facility to produce utilities more efficiently in the long-term, a Net Analysis should be used to model both the losses to private Industry and the cost savings to Government (increased spending) or Households (additional HH income).  This assumes, however, that the utility company would not continue to produce that Commodity and that the previous local consumption could not be replaced by non-local consumption (exports).

RELATED TOPICS:

Electricity Generation + Distribution FAQ

Methodology for Development of the 2017 Detailed Production Functions for IMPLAN’s Nine Electrical Power Sectors

INTRODUCTION:

Spending Patterns can be thought of as the backward linkages or supply chain of the Industry. Spending Pattern Events are best used when modifications to an Industry’s Production Function are necessary that cannot be made with a standard Industry Event. When detailed information is known about the Industry’s spending, a Spending Pattern Event can be used and adjusted to reflect the specific purchases or ratios of purchases. Spending Patterns include all Intermediate Expenditures for a given Industry.

One size may fit most, but maybe not all. While the Spending Patterns represent the average buying pattern and ratios of all firms in that Industry, it may not match the expenditure pattern of your firm. For example, if you are working with an organic farmer, you might note that there are no Sectors for organic farming.  As a result your Indirect Effects include a large quantity of pesticides and herbicides that just aren’t appropriate. You can use the Industry Spending Pattern to adjust the purchases of your firm with your detailed knowledge of their spending.

DETAILS:

The default in IMPLAN is set the Event Value of an Industry Spending Pattern to Intermediate Expenditures. This includes the purchases of non-durable goods and services such as energy, materials, and purchased services that are used for the production of other goods and services (rather than for final consumption). When Intermediate Expenditures is chosen, the Event Value will be run through the model.

However, only total Output may be known.  Output is the total value of Industry production in producer prices.

• Manufacturing Sectors: Output = sales plus/minus change in inventory
• Service Sectors: Output = sales
• Retail & Wholesale Sectors: Output = gross margin (or Marginal Revenue) 

Remember: 

When Output is chosen, the value will be multiplied by the total Gross Absorption percentage for the Industry before being run through the model.

INDUSTRY SPENDING PATTERN:

Industry Spending Patterns allow you to build an Industry from data about its expenditures. The coefficients listed in an Industry Spending Pattern Event represent the amount spent on each commodity per dollar of that Industry’s Intermediate Expenditures.

An important note about these spending patterns is that the direct effect is not accounted for in the results.

Let’s say we want to model the Industry Spending Pattern for a new fast food restaurant (Sector 502 – Limited service restaurants) that is planning to open up in Rochester, NY (Monroe County).  Opening up the Advanced Menu of our Event, we see the complete list of Commodities purchased by fast food restaurants in our Region.

We know that our restaurant, Stephanie’s Subs, has projected Sales of $5M.  We also know that they will not be serving alcohol, so we want to eliminate this spending (Breweries, Wineries, and Distilleries).  We can scroll down to find Commodities 3108, 3109, and 3110 and change the given percentage to 0% in each of them.  

After our changes, we can see that the total percentage now sums to 97.18%.  Clicking on the Advanced Menu allows us to Normalize the results so that they once again sum to 100%.  Also, this is where we can reset the Industry Spending Pattern to IMPLAN defaults.  

We also have the option to change the Local Purchase Percentage (LPP) inside the Industry Spending Pattern.  The default in IMPLAN is that the LPP is set to the Social Accounting Matrix (SAM) value. Unless you know that the Commodity will be purchased locally, it is advised to leave the SAM value.

Now we are ready to hit Run and examine our Results.

Notice that there are no Direct Effects in our Results so we will need to add those back in manually.  We know our total Output is $5M. Now we need to navigate back to our Regions screen and click on 

Study Area Data > 

     Regions Industry Summary

Here we can find IMPLAN’s known totals for Sector 502 – Limited service restaurants for Output, Employment, Labor Income, and Value Added.  To model the Induced Effects from the Labor Income, Analysis-by-Parts (ABP) would be the best method.

Calculating the ratio of Stephanie’s Output to the total Output in the county, we see her restaurant will be 0.43% of the total.  We can then apply this percentage to the other values in our impact to estimate the Direct effect. The template for these calculations is here. 

EXAMPLES:

Editing Institutional Spending Pattern Events

ABP: Analysis-by-Parts Using an Industry Spending Pattern Event with Labor Income Event(s)

Hospitals: Modeling Private Hospital Impacts with Analysis-by-Parts

INTRODUCTION:

More and more people are interested in the economic impact of tourism – how guests in a region spend their money when visiting.  The World Travel & Tourism Council estimates that one in ten jobs are related to tourism, so it’s no wonder that governments and economic development organizations are interested in the local and regional effects.  Public money is invested not only in infrastructure and improvements to encourage tourism, but also on large marketing campaigns domestically and internationally.  

Tourism is of great interest to many regions of the country. In impact analysis, tourism is unique among export activities. Exports are typically thought of as goods or services shipped by local businesses to purchasers located outside the region. Tourism, on the other hand, involves purchasers traveling into the region to acquire goods and services from local businesses. Tourism is most often classified by the activities in which non-locals participate: film festivals, fall colors, professional sporting events, historical sites, conventions, and so forth. Lodging, restaurants, retail, and recreational services typically rely upon this kind of export as key to their success. Because tourism comprises a wide variety of activities and types of businesses, there is no single “Tourism” sector for modeling impacts.

Stynes (n.d.) poses just a few of the questions that an economic impact analysis of tourism spending can help the analyst answer:

• How much do tourists spend in the area? 
• What portion of sales by local businesses is due to tourism? 
• How much income does tourism generate for households and businesses in the area
• How many jobs in the area does tourism support?
• How much tax revenue is generated from tourism?

KEY CONSIDERATIONS:

DEFINING TOURIST

For an impact on the state, anyone traveling from outside would be considered a tourist.  Why do we only include non-local tourists in the analysis? Sales to final demand are the classic activities modeled by input-output analysis. These sales (direct effects) result in new production of goods and services that initiate a chain of economic events in the region (indirect and induced effects). While local residents may spend money at these tourist attractions, their expenditures are not typically counted in tourist studies because if those tourist attractions did not exist, the residents would likely still spend that money locally.  If they didn’t go to the festival, they might have spent their money bowling or at the movies. That said, a case can sometimes be made for including resident spending on tourist attractions if the study area is very small and has few alternatives for recreation, such that the existing tourist attractions keep residents’ money from leaking to other places (i.e., they keep residents from traveling elsewhere for their recreational pursuits). This is known as import substitution – someone doesn’t have to go to Nevada to gamble if their state legalizes the industry – so the money is no longer leaked out of the local economy.

TOURIST TYPES

There are many different ways to segment tourists.  For example, a single individual traveling for the biannual MCRSA/IMPLAN Conference will spend money very differently than a family heading to camp in a National Park. Also, domestic tourists and international tourists are also likely to spend very differently.  Consider segmenting your groups to model different kinds of tourists separately. In general, keep these differences in mind when collecting data or analyzing data given to you.  Some helpful tips are available in the article about Surveying for Input-Output.  

HOW TOURISTS SPEND

Based on the parameters of the tourist survey or the source of our tourist spending data, we may not know if the responses reflect the same purchasing patterns as our local tourists, or if responses were constrained to only describe local purchases. In this case, Local Purchase Percentage (LPP) could be customized to ensure our impacts were conservative and to reflect the idea that tourist purchases in the region follow a pattern like resident purchases regarding locality.

TEMPORARY SPENDING

Tourist spending is likely somewhat more volatile than operational spending.  Although it is projected to grow, influences like consumer preferences, generational differences, political climate, and economic conditions are changing the way people travel (Tour Writer, n.d.).  Keep these considerations in mind when using older survey data or trying to predict too far into the future.

PREPARING DATA FOR IMPLAN:

Very simply, a study of tourist spending looks at how visitors spend money in a local economy. We are either given data on tourism spending by or must survey visitors to obtain a breakdown of their spending.  

Let’s say we want to look at the economic impact of 100 people visiting an amusement park in California for the day.  We are given survey results that show each person spends an average of $50 on gasoline, $25 on snacks, $150 in restaurants, and $200 on amusement park tickets, and $75 on retail merchandise, for a total of $500.  We can model these through an analysis by assigning one of the IMPLAN Sectors to each of the spending categories and inputting the totals. Since we know what items people were purchasing, but not necessarily from what businesses they bought, we use Commodities.  

LOCAL PURCHASE PERCENTAGE

For the gasoline, we can only assume that about half of that money would stay in the state economy; as some portion is likely leaking to other states on their road trip.  To adjust for this, set the Local Purchase Percentage (LPP) to 50%.  In our example, the table above is all local spending expect for the gasoline, so for each of our Commodity Events,, LPP will be left at 100%. 

MARGINS

For the retail spending, we asked the guests what they spent while in town for the event, so we need to think about margins. The prices that were paid at stores represent Purchaser Prices or Total Revenue, so we need to check the box for Total Revenue to tell the model that to convert these to Producer Prices or Marginal Revenue before applying the Multipliers.

Note that for gasoline we will use the producing Commodity (3156 – Refined petroleum products) even though there is a retail gas station sector in IMPLAN. The reason for this is that we know exactly what was purchased from the retailer in this case (gasoline and oil) and can thus identify the producing sector (Refined petroleum products), whereas in the other retail spending categories we don’t know exactly what was purchased and thus cannot identify the producing sector. For example, we don’t know what Retail consisted of and therefore do not know how to choose between the various manufacturing sectors (tshirts, mouse ears, postcards, etc.). Thus, in this latter case the best we can do is capture the retail margin portion of the purchase, whereas in the case of gas and oil we can also capture some of the producer value (if there is a local producer), as well as the transportation and wholesale margins, in addition to the retail margin. While you could use the retail gas station Commodity (3402 – Retail services – Gasoline stores), you would only capture the retail margin.

DEFLATORS

Sometimes we need to rely on older survey data for the analysis.  We might have a survey that reflects tourist spending in 2009, but we are looking to estimate the impact of tourists in our region in 2019. Since the survey is vintage 2009, the Event Year should reflect that the entered dollar values are 2009 dollars (i.e. Event Year = 2009). If we want to see the economic impact of tourist spending in 2019, we can use the Dollar Year for View on the Results page to adjust the values to 2019 dollars.

Deflators are specific to the dollar value of the investment or expenditures, it is important to ensure that the Dollar Year reflects the dollar value being entered. Because the model’s relationships are all based on the year of the data set or the Data Year (for this example, 2017), dollar values for years other than 2017 must be adjusted to 2017’s prior to those values being applied to the multipliers.  This conversion is done by IMPLAN’s built-in deflators. Deflators are units of measure, not descriptions of when an Event is taking place.

THE PROCESS: 

Now we are ready to put the data in IMPLAN.  Create the California Region and add each line of spending as its own Commodity Output Event.  In our example, we have the per person spending, so we enter that in the Event line. Then on the right side of the screen, we use the Scaling option to reflect the 100 visitors.  Alternatively, we could multiply each of the Commodity totals by the 100 visitors to get the same result.

THE RESULTS: 

In this example, we had 100 visitors to California, each spending $500.  We used data from 2017 and are viewing the results in 2019 dollars. We also put in $75 per person in Sector 3405 – Retail services – General merchandise stores, which only captures the retail margin.  Therefore, we see the total Direct Output as slightly less than the $50,000 in spending.  

We can also see the large Direct Output Impact in Sector 494 – Amusement parks and arcades and Sector 501 – Full-service restaurants, which is to be expected from the large per person spending in these categories.  

TIPS FROM THE WEB:

Neffke, F., Ravinutala, S., & Zuccolo, B. (n.d.).  Tourist Spending Insights Provide Unprecedented View of Global Tourism. Retrieved from 

https://www.hks.harvard.edu/centers/cid/about-cid/news-announcements/tourist-spending-insights.

Stynes, D. (n.d.). Economic Impacts of Tourism. Retrieved from https://msu.edu/course/prr/840/econimpact/pdf/ecimpvol1.pdf.Tour Writer. (n.d.). Tourism industry statistics for 2018 and beyond.  Retrieved from https://www.tourwriter.com/travel-software-blog/2018-tourism-stats/

US Travel. (2019).  U.S. Travel and Tourism Overview (2018).  Retrieved from https://www.ustravel.org/system/files/media_root/document/Research_Fact-Sheet_US-Travel-and-Tourism-Overview.pdf.  

World Travel & Tourism Council. (2019).  Travel & Tourism Economic Impact. Retrieved from https://www.wttc.org/economic-impact/.

INTRODUCTION:

Sometimes the Industry that we want to examine does not exist in our study Region. The Industry might be new to the area.  It also might be a new Industry altogether. It also might be an industry that only recently became legal. This article outlines the best way to model these new Industries in IMPLAN.  

DETAILS:

The economy is always changing and Industries come and go.  Federal data follows these changes by adjusting the North American Industry Classification System (NAICS) codes.  IMPLAN follows suit, so when the NAICS codes change, so do IMPLAN Sectors.

There are two ways to create a new Industry depending on what you want to accomplish.  Both methods are outlined below.  

INDUSTRY THAT EXISTS IN THE US:

If you have a Region that will be gaining an Industry that does exist somewhere in the US, you can follow the steps outlined in the article Adding an Industry by Customizing a Region.

BRAND NEW INDUSTRY:

Federal data lags in categorizing these new Industries, which can pose a problem for researchers.  When the newest widget is invented, it is usually categorized under with a similar product or under one of the catch-all codes like “all other…”  One interesting case is that of the legalization of cannabis production and sales. Canada has already outlined the NAICS codes for the entirety of the cannabis Industry: Classifying Cannabis in the Canadian Statistical System.  

Depending on what part of the supply chain we want to examine, we find the most similar NAICS or IMPLAN code and edit the Industry Spending Pattern for that Industry.  In the cannabis example, we may want to model the agricultural impact using one of these Industries: 2 – Grain farming; 6 – Greenhouse, nursery, and floriculture production; or 7 – Tobacco farming.  The manufacturing might fall under Sectors 111 – Tobacco product manufacturing or 173 – Medicinal and botanical manufacturing. Finally, the retail component will fall under 406 – Retail – Miscellaneous store retailers.  

Let’s say we want to examine the effect of a $100M in cannabis agricultural production in Ohio and we know that a full $5M of that will be spent on legal fees. We also know there will be 25 employees each making $50,000.

First we have to decide what Sector we want to use.  In this case knowing the climate in Ohio, we make the assumption that the agriculture will be most closely related to IMPLAN Sector 6 – Greenhouse, nursery, and floriculture production.  

Create an Industry Spending Pattern Event for $100M in Sector 6.  If we know more details about the spending of cannabis growers as compared to other greenhouse products, we can edit the spending pattern to further reflect this new Industry.  Remember in this example we know that $5M will be spent on legal fees specifically. To model this, we click the Advanced button to open the details of the Industry Spending Pattern and scroll down to Commodity 3447 – Legal Services.  Next, we override the value with 5%; the dark blue color shows the edited Commodity. If any further information is known including Sectors that could be deleted entirely from the Spending Pattern, that can be done at this point as well. More details on this can be found in the articles Editing Industry Spending Pattern Events or Editing Institutional Spending Pattern Events.  

You’ll notice that the Sum of Percentages at the bottom is now 104.9%.  

To fix this and return to 100%, click on the Advanced button and choose Normalize.

Now, a full 5% will be spent on legal fees.  Make sure to choose Total Output (instead of Intermediate Expenditures). Remember: Output = Intermediate Expenditures + Value Added.

Next, we create a second Event for our Labor Income (25 employees x $50,000 = $1,125,000).  We can then add them to our Group and Run our economic impact.

Notice that there are no Direct Effects in our Results so we will need to add those back in manually.  We know our total Output is $100M and our Labor Income is $1,125,000 with 25 employees. Now we need to navigate back to our Regions screen and click 

Study Area Data > 
     Regions Industry Summary

Here we can find IMPLAN’s known totals for Sector 6 – Greenhouse, nursery, and floriculture production. Calculating the ratio of the overall Value Added to the overall Output shows that VA is, on average, 65% of the total Output.  Therefore, we multiply $100M in Output by 65% to yield a Direct Value Added for this project of $65,620,927. The template for these calculations is here.

RELATED TOPICS:

Adding an Industry by Customizing a Region

Editing Industry Spending Pattern Events

Editing Institutional Spending Pattern Events