by Paul Losleben, Orcas Research Group
Objective
What happens to a dollar spent on Orcas Island? Portions of that dollar are re-circulated in the form of salaries and business profits which are, in turn, spent in part again on the island. Other portions leave the island in the purchase of goods and services from the mainland. The extent to which money re-circulates and thereby stimulates the local economy is called a multiplier effect. Specifically, we define multiplier to be the additional spending power of the origional dollar that is created in a local economy. For example, if one dollar spent in the local economy produces 80 cents of additional local spending, the multiplier is .8. The objective of this study is to determine the multiplier effect for Orcas Island.Background
Multiplier effects are often used by economists in economic impact studies, for example, to help justify a new shopping center, sports stadium or public economic policy. Unfortunately, these studies often overstate the benefit of some new proposed project in order to justify public concessions[1]. Locally, for example, in the interest of encouraging local residents to shop locally it was widely reported that a dollar spent on Orcas Island re-circulates 7 times[2]. Unfortunately, multipliers for rural communities tend to be much lower[3]. In addition, multipliers for a small region like Orcas Island tend to be smaller than a larger geographic area like a state or nation. Statistics of this type are often accepted without challenge since it is difficult to independently develop or verify this kind of information. The studies often cost in the hundreds of thousands of dollars. This study takes advantage of what is a fairly simple economy on Orcas Island. There is essentially no manufacturing and very little business to business commerce. The vast majority of financial transactions are found in the purchase of goods and services by the residents for their personal use. With this simplification, (and estimates of the error introduced) the following methodology was devised.Methodology
Let P be the net percentage of a dollar that is re-circulated and let M be the total amount of additional local spending that is produced as a result of the remainder of each dollar which is, in turn, spent again and again in the local economy. . It might be useful to think of this effect as if the dollar were a bouncing ball. Each time it bounces some amount of energy is lost and the remaining energy contributes to the next bounce. The total amount of energy that contributes to all the bounces is the multiplier. If we have a really good ball, it will bounce lots of times and the multiplier will be quite large. On the other hand, if it is a tennis ball that the dogs have been playing with for a while, it may only bounce a few times and the multiplier will be quite small. Just as we would like to have new tennis balls when we go to the court to play tennis, so also we would like our economy to have lots of bounce. Unfortunately, we sometimes find that it has gone to the dogs!
Or
We see from the chart that as P increases (to the right of the graph), M increases rapidly and approaches infinity as the P approaches 100%. Likewise, as P decreases (to the left of the graph), M approaches zero. For P = 50%, M is exactly unity or one dollar spent in the local economy will produce exactly one additional dollar. The shape of the right side of the curve should be fairly intuitive—the less money one removes from an economy, the more times money will re-circulate, until at the limit, if no money is removed, money will re-circulate infinitely What is perhaps not so intuitive is that the curve is relatively flat for values of P less than 50%.
From the shape of the curve, it should be cautioned that conclusions based on larger values of P are highly subject to error. For example, for P = 90%, a ±1% error would produce an error of over ±10% in M! The message here is to beware predictions of high multiplier values. On the other hand, for values of P less than 50%, the error in M will become progressively smaller than 1:1 as P becomes smaller.
This study uses consumer statistics to determine the amount of money that is re-circulated for each category of a typical family budget. The overall percentage P of a dollar that is re-circulated is represented by:
Where Bi is the percentage of personal income spent for a budget category and Pi is the percentage of dollars spent in that category which is re-circulated.
Detail
Assumptions
The primary source of consumer statistics for this study is data published by the United States Department of Labor, Bureau of Labor Statistics (BLS)[4]. The data used is for the Seattle-Tacoma-Bremerton area and, as such, cannot be assumed to be absolutely accurate for Orcas Island. Since similar data is not available for San Juan County, much less Orcas Island, we shall use this data and discuss the probable accuracy of the data for this study where applicable. The percent distribution by category shown in Chart 2 is derived from Table 1 in this report.
Federal Taxes 15%
The amount of money which is returned to the county by the federal government includes Social Security benefits, Supplemental Security benefits, military and federal government pensions, various federal grants, contracts and salaries. By far, the largest contribution is Social Security ($26.9M). Altogether, $55.1M of our federal taxes in 2000 were returned to San Juan County, primarily in the form of Social Security, Supplemental Security and other federal retirement benefits[5]. We estimate this to be 80% of the federal taxes paid that year, a figure which is surprisingly large.B1P1 = .15 * .8 = .12
Food and Beverages 14%
National figures for supermarkets indicate that labor constitutes 11% of sales[6] and net profit after taxes is 1.36% of sales[7]. These figures are primarily based on large supermarket chains and it is useful to examine similar figures for the smaller store sizes typical of the islands. By contrast, cooperative groceries[8] with annual sales under $750K report labor costs at 20% of sales and net income at 2.8%. Those with annual sales up to $2M report labor costs at 23% and net income near 0%. Taking a conservative estimate, we will assume for a typical island-owned market that 23% of sales remains on the island in the form of wages and profit[9].Not all food and beverage expenditures are made in the purchase of food from markets. Prepared foods also need to be considered. As one might expect, statistics for the restaurant industry depend heavily on many factors involving the type of restaurant. Some restaurants pay a premium for the food that they purchase while others emphasize labor reduction and efficient service. We will use figures reported by McKinsey[10] for a profitable business with moderate labor turnover reporting labor costs of 34% and profit of 10%. This yields a figure of 44% of sales retained on the islands for locally owned restaurants.
For the Seattle area, the BLS statistics break out food consumed at home (7.5%), food consumed away from home (5.3%) and alcoholic beverages (.8%). Realistically speaking, residents of the islands do not eat out as frequently as their counterparts in Seattle. Without proof, we will assert that a conservative estimate would place food consumed at home at 12% and food consumed away from home at 2% of the total budget or an amount equivalent to eating out once per week.
B2P2 = .12 * .23 + .02 * .44 = .036
Housing 29%
The BLS figures break out five subcategories of housing which we shall investigate separately: Shelter (18.5%), Utilities (4.3%), Household Operations (1.3%), Household Supplies (1.2%) and Household Furnishings & Equipment (3.2%).The cost of shelter includes rent or mortgage payment and real-estate taxes if applicable. For the purpose of this analysis, we assume that housing is built with local labor and materials that are purchased from local building supply companies. Mortgages are almost always held by off-island businesses (Islanders Bank is the only financial institution which retains a local portfolio of variable rate mortgages)[11]. Consequently the interest contribution is the most significant portion of shelter cost and nearly all leaves the island. Assuming a 30 year mortgage at 6%, interest constitutes 54% of mortgage payments. Real estate taxes comprise approximately .8% of assessed value. If we account for the amount of real estate tax that is returned to the county, approximately 50% of all payments for basic shelter leave the island.
The cost of utilities includes electricity, phone, gas, water and sewer where applicable. The cost of energy, primarily for heating dominates these figures. Referencing the 2000 census figures[12], over half of the homes in San Juan County are heated by electricity. 28% are heated by wood, the only energy source produced on the island. The local power company purchases electricity from the mainland and contributes 20% of it's income to the local economy, primarily in the form of wages. Phone, heating oil and propane have a much lower percentage payroll. Conversely, local water companies purchase supplies from the mainland, but are owned and operated locally[13]. Sewage is almost entirely on private septic systems. Overall, we estimate that 25% of all utility costs remain in the local economy.
The cost of household operations includes upkeep and maintenance. For the purpose of this analysis, we will assume that local sources are used for all upkeep and maintenance costs.
The cost of household supplies includes various cleaners, paper supplies, etc. For the purpose of this analysis, we will assume that all household supplies are purchased locally.
The cost of furnishings and equipment includes furniture and appliances. For the purpose of this analysis, we will assume that all furnishings and equipment are purchased off-island.
B3P3 = .185 * .5 + .043 * .25 + .013 + .012 + .032 * 0 = .13
Apparel and services 4%
Nearly all clothing is purchased off island and laundry services are nearly zero. For the purpose of this analysis, we will count this category as having negligible contribution to the local economy.B4P4 = 0
Transportation 14%
The BLS figures break out four subcategories of transportation that we shall investigate separately. Vehicle purchase (5.4%), Gasoline and Oil (2.5%), Other Vehicle Expenses (5.2%), and Public Transportation (1.3%).There are no auto dealers in San Juan County and for the purpose of this analysis, we will assume that all nearly all autos are purchased off-island.
Fuel in the form of gas and diesel are imported from the mainland and for the purpose of this analysis, we assume that only 10% of these costs are returned to the local economy in the form of wages and profit.
Other vehicle expenses include insurance and maintenance. For the purpose of this analysis, we assume that 50% of these costs are insurance purchased from off-island companies and that the remaining 50% is equally divided with half of all vehicle maintenance performed locally.
Public transportation in the form of taxi and bus service is minimal. For the purpose of this analysis, we assume that 90% of these costs are in ferry service (allowing for the salary paid to local dock workers) and therefore do not contribute directly to the local economy.
B5P5 = .054 * 0 + .025 * .1 + .052 * .25 + .013 * .1 = .017
Health care 5%
Health care includes both the salaries paid to the local health care providers and the off-island expenses of laboratory expenses, hospital care and administration. For the purpose of this analysis, we will assume that 50% of health care expenses are retained in the local economy.B6P6 = .05 * .5 = .025
Entertainment 4%
Entertainment is available on Orcas Island in a variety of forms ranging from video rental to live performances at the Orcas Center. Depending on the type of entertainment, the share of the cost that is sent to the mainland varies widely. For the purpose of this analysis, we will assume that 90% is retained in the local economy.B7P7 = .04 * .9 = .036
Contributions 2%
The residents of Orcas Island pride themselves in their generosity and many public services exist by virtue of donations. For the purpose of this analysis, we will assume that all contributions are made to local organizations.B8P8 = .02
Personal insurance and pensions 6%
Insurance premiums are generally paid to institutions on the mainland and do not contribute directly to the local economy. In a different way, pensions do not count as income until actually claimed. While money invested into pension accounts increases the personal wealth of island residents, there is little in this category which counts as being returned immediately to the local economy.B9P9 = 0
Other 5%
For the purpose of this analysis, we will assume that half this category is retained in the local economy.B10P10 =.025
Summary
Combining the above approximations: P = .41 and therefore 
or 67 cents of additional local spending is produced for every dollar spent locally.
Error Analysis
The BLS data used for the above analysis are developed for the Northwest United States region, specifically based on surveys taken in the Seattle area. How well does the data match San Juan County? Chart 3 shows the year 2000 census data[14] for distribution of income by households for San Juan County as well as the three counties corresponding to Seattle, Tacoma and Bremerton which are used in the BLS data. The distribution of earnings is quite similar for all four counties, although with a slightly lower median household income for San Juan County ($43,491) in comparison to King County ($53,157), Pierce County ($45,204) and Kitsap County ($46,840).
Does this make a significant difference in spending patterns? Do wealthier households spend more as a percentage of their income than poorer households? Returning to the Bureau of Labor Statistics[15], we derive the following for national averages:
As a percentage of disposable income, wealthy households spend more on Taxes, less on Food, less on Housing, the same on Apparel, less on Transportation, the same on Health Care, less on Entertainment, the same on Cash Contributions, more on Insurance and Pensions, and the same on Misc. We use the figures derived from this report to develop an estimate of approximate error in each of these categories. The most significant differences are, not surprisingly, Taxes, Housing, and Insurance & Pensions.
Combining the above approximations where P is now in the range of 40% to 45%:
One might argue that there are also unique aspects of the island culture, transportation costs, and demographics which would also introduce error. For example, it is difficult to believe that island residents spend as much on apparel as their counterparts on the mainland when "Orcas Casual" implies denim and fleece. In the absence of detailed statistics for several of the above assumptions, we assert that this represents an upper bound to the multiplier. In each case where we assumed that a local merchant or tradesman was used, we used a conservative estimate. In actuality, we believe that many more purchases are made on the mainland. For example, some percentage of groceries and other household supplies are purchased at COSTCO, island houses are increasingly factory-built, many of us use off-island medical care, etc. One reviewer of this study estimates that the figures for health care and entertainment are high by a factor of 2x to 3x.
As the reader may observe from the multiplier curve, these differences do not introduce significant error. It is useful to note where this data resides on the multiplier curve. As the percentage re-circulated decreases, the curve becomes increasingly flat. This means that values lower than those derived, which would represent decreased revenue retained on the island, become less sensitive to error. Therefore, we claim that P is in the range of 40% to 45% and the multiplier is at most .82 and is possibly lower than .67.
Finally, this analysis so far has been based solely on how money is spent in a typical family budget. This assumption is valid for all terms after the first term in the series where money ultimately recirculates in the local economy primarily in the categories described above, but is not valid for the first term if the initial dollar is spent by a non-resident. For example, a dollar spent in the local economy by a non-resident counts only in the category where money is actually spent for the first term of the series. There is no portion of that dollar that counts toward federal income taxes, automobile purchase, home ownership, or any other category other than that where the money is actually spent. It is only when the money reappears as wages and profits that it counts in the second and all subsequent terms in the series. How seriously does this change the results? Only 29% of expenditures by all travelers to the San Juan Islands result in wages that remain on the island[16]. Even considering profits, this is less than the initial term in the series for residents (41%) and consequently if it has any effect, it will reduce the multiplier by a small amount.
Conclusions
We conclude that money leaves the island very quickly. Simply stated, a dollar spent locally produces at most 82 cents of additional local spending. Furthermore, it is quite possible that less than 67 cents of additional local spending is generated.
What does this mean? To first order, the multiplier is a measure of the self-sufficiency of a community. This analysis demonstrates the extent to which we are dependent on a continuing flow of money from the mainland. Simply stated, we produce very little in comparison to what we consume. Instead, we are heavily dependent on a flow of money that can be attracted to the island. For Orcas Island, this consists primarily of retirement and investment income (nearly 50% of reported income in the 1997 economic census) and, to a lesser extent, on businesses related to construction and tourism. As an increasing population also increases demand for land and services, growth in these areas of the economy increases land assessments and raises taxes to pay for new infrastructure. Meanwhile salaries in the service sector, in construction and tourism have remained low. As a result, the working middle class is increasingly finding it difficult to make ends meet.
There are three approaches to improving the situation:
First, we might simply encourage our residents to shop locally. While it is beyond the scope of this analysis, it could be argued that it is false economy to shop for bargains on the mainland since the cost of transportation and personal time often exceed any savings. Even so, it is difficult to convince our residents to ignore savings at high volume outlets like COSTCO or LOEWS when viewed as part of a planned trip to the mainland for other reasons. Moreover, many products and services are simply not available on the island. The main point to be learned here is that only marginal improvements in the multiplier can be expected by simply encouraging our population to shop locally. Rather, we value the convenience and personal service in shopping locally and we need to patronize our local merchants if we want them to survive.
Second, we might become more self-sufficient by encouraging the establishment of businesses that retain dollars on the island. For example, a local source for fixed rate mortgages would be beneficial since the interest would remain on the island. Marketing of home-grown produce is another good example. Unfortunately, there are few good business cases that can compete with the larger scale of operations that most businesses on the mainland enjoy. The trend toward increased use of electronic commerce on the Internet is demonstrating for many businesses that geography doesn't matter. A sufficiently large economy of scale can be achieved so that these new, remote businesses can compete, even with increased shipping costs. As a result, many residents now shop on the Internet and that trend can be expected to increase.
Finally, the most valuable thing we could do would be to encourage the establishment of businesses that generate the majority of their revenue by selling products or services to customers off-island. This approach, of course, does not change the multiplier, but seeks to simply diversify and increase the sources of money from off-island. There is one notable case where the multiplier is increased. If island residents invest in island businesses as part of their investment or retirement plan, the multiplier is improved in both the tax and pension categories. We believe that it is useful, therefore, to encourage the establishment of small businesses serving off-island markets and preferably those which are financed locally. We will pursue this further in future work.
The author of this study welcomes constructive criticism of this work. Please address correspondence to: losleben@rockisland.com.
References and Footnotes
[1] Olson, Mancur, The Logic of Collective Action, revised, Cambridge: Harvard University Press, 1971.[2]This number, besides being confusing, was traced by the author to the well-intended efforts of a member of the local Chamber of Commerce who got it from the Friday Harbor Chamber of Commerce. They got it from the Tahoe, CA Chamber of Commerce, who got it from another Chamber in California. It is a good example of a statistic that people would like to believe because if it were true it would encourage support of local business, a commendable if somewhat misinformed motivation.
[3]Burress, David; Economic Impact Multipliers for Kansas; "Kansas Business Review" Vol. 12, No. 3, Spring 1989
[4] "Consumer Spending Patterns in Seattle-Tacoma-Bremerton, 1999-2000," U.S. Department of Labor, Bureau of Labor Statistics, Document 9780
[5] Consolidated Federal Funds Report, San Juan County, Washington State, U.S. Census Bureau, Year 2000.
[6] Supermarket Operating Costs, 2001, Key Industry Facts, Food Marketing Institute Information Services, June 2002.
[7] Food Retailers' Net Profit—Percent of Sales, Key Industry Facts, Food Marketing Institute Information Services, November 2002.
[8] W. Swanson and D. Gutknecht, Retail Operations Survey, Cooperative Grocer Online, Report # 89, July-August 2000.
[9] The figure of 23% was verified as typical of Island Market in private correspondence with Dale Linnes, February, 2003.
[10] David S. Friedman, "Help wanted," The McKinsey Quarterly, 1998 Number 1, pp.34-44.
[11] Private conversation with Jerry Wilson, Islanders Bank, February 3003.
[12] Table DP-4, Profile of Selected Housing Characteristics: 2000, U.S. Department of the Census.
[13] We note that the Rosario water company is owned off-island and managed locally.
[14] Table DP-3, Profile of Selected Economic Characteristics: 2000, U.S. Department of the Census.
[15] "Consumer Expenditures in 2000," Table 2, U.S. Department of Labor, Bureau of Labor Statistics, Report 958
[16] "Washington State County Travel Impact 1999-2001," October, 2002, Dean Runyan Associates, Portland, OR.