The underpinning theory of IMPLAN

Comments

12 comments

  • Avatar
    IMPLAN Forum
    Hello Kei That is an excellent question, but one that is difficult to summarize briefly. In general, IMPLAN uses a Social Accounting Matrix model, which is an Input-Output model that is expanded to include inter-institutional transfers (e.g. transfers among households and government). Because the SAM contains all the information as to the inputs required by each industry (both labor and non-labor) and the taxes paid by each industry, as well as information on household spending and taxes and information on local vs. non-local purchasing, SAM models can predict the ripple effects (i.e., the effects on other industries, government revenues, etc.) of a change in one or more specific industries, or a change in household spending, etc. based on the assumption that the indirectly-affected industries and households will respond by spending the same way they did in the year of the IMPLAN data set (i.e., as reflected by the underlying study area data). Major assumptions of I-O analysis include: 1. Fixed production technology (no input substitution). To increase output by 10%, an industry needs 10% more of every input 2. No supply constraints. If an industry needs an input, it can get it (whether it is imported or purchased locally IS accounted for in IMPLAN via the RPCs). It's also important to note that while Version 3 of IMPLAN provides a framework to conduct an analysis of economic impacts, and that framework embodies certain assumptions and theory, how you use Version 3 will bring additional assumptions and theory of your own to the analysis you conduct using IMPLAN. To understand more details of the theory underpinning IMPLAN, we recommend that you consult articles posted to our website [url=http://implan.com/index.php?option=com_content&view=article&id=821%3Aresearching-implan-data&catid=185%3Adata-information&Itemid=166]here[/url]. Articles on that page summarize the data collection method, key assumptions about Input-Output analysis and IMPLAN, and data sources. An additional article describing an overview of the data collection process can be found [url=http://implan.com/index.php?option=com_content&view=article&id=689%3Aimplan-data-source-outline&catid=253%3AKB33&Itemid=166]here[/url]. Please let us know if you have any follow-up questions.
    0
    Comment actions Permalink
  • Avatar
    Kei Wei
    Thank you for your explanation.Do you mean that there is no theory underpinning the IMPLAN? It is just the adaption of IO model and SAM? I purchased four data on Maui, Hawaii, Honolulu and Kauai but I am not sure which county data is similar and suitable to my study area since I am not familar with US. My study area is a small area 442km2, population 133,164. The main economic activities are agricultural and tourism. This place is famous for diving and holiday.How I can modify the US data into my case in Malaysia since I do not know the local taxes, household spending and tourism employment? The only data I collected were the expenditure by visitors using the IMPLAN sectors.IMPLAN generates results on employment, income and output, how the information truely reflect the local situation since the data is from US and I am doing it outside US? I am sure this is important to me to justify.
    0
    Comment actions Permalink
  • Avatar
    IMPLAN Forum
    Hi Kei, We aren't sure exactly what you are looking for in regards to the underpinning theory, but here is the mathematical basis of the underlying Multipliers Model. Leontief, Wassily. The Structure of American Economy, 1919-1929; an Empirical Application of Equilibrium Analysis. Cambridge, Mass.,: Harvard University Press, 1941. The above citation is the tap root to the theory of input-output analysis. If output equals x and factor shares equal A and final demand equals d, where lower case letters are vector and upper case are matrices, then x=Ax+d solving for x (I-A)x=d x=(I-A)^-1*d A change in final demand (d) will give you a measure of the change in output (x) across all sectors as reconciled through the multiplier matrix (I-A)^-1. This is a simple but powerful insight into a complex economic system. Hopefully that addresses your concern. If not could you provide some additional details as to what exactly you are trying to determine in regards to the underpinning theory. As regards choosing a Model, you can build the 4 HI counties and view the Study Area Data via the Explore menu. The very first view that opens is the Industry Detail view, from this you can discern which county appears most like the regional economy you are looking at working with. In addition to this, we would recommend doing some digging to determine your local Employment Compensation per worker, Output per worker, etc. and how this compares between the region you area and the Hawaii county you select, and make any adjustments to the Model as needed. If you don't have any better data with which to modify the values, the best you can do is use what is provided in the county, and explain that this was done because you had no better data available to you. You can compare items like population on the basis of the Model Overview. Regarding tax impacts, you will have again need to try to see if you can find comparable data for your region, or justify the use of the HI values on the basis of the fact that no better data was available to you. The impacts of the Sectors you select will provide you the Employment relative to the Employment resulting from impact study, but as regards base Employment unless you have values with which you can modify the base-line in IMPLAN or know of governmental sources you could reach out to, to try to get that information, we unfortunately we don't have any other sources with which to assist.
    0
    Comment actions Permalink
  • Avatar
    Kei Wei
    I do not have the local Employment Compensation per worker, Output per worker as it is not available in my study region. Also, i do not have the tax information at local level. If I use the HI data, i might be questioned on the reliability of my result. I guess most of the local people in HI are engaging in tourism related sectors and they are not as poor as the local people in my study area. Most of the local people in my study area work as fishermen and earn less than USD300 per month. How can I make justification on it since it looks like I am comparing orange and apple?
    0
    Comment actions Permalink
  • Avatar
    Kei Wei
    I do not have the local Employment Compensation per worker, Output per worker as it is not available in my study region. Also, i do not have the tax information at local level. If I use the HI data, i might be questioned on the reliability of my result. I guess most of the local people in HI are engaging in tourism related sectors and they are not as poor as the local people in my study area. Most of the local people in my study area work as fishermen and earn less than USD300 per month. How can I make justification on it since it looks like I am comparing orange and apple?
    0
    Comment actions Permalink
  • Avatar
    IMPLAN Forum
    Hi Kei, Given that you do not have the local Employment Compensation per worker, Output per worker for your study region, as stated in an earlier Post, you could use the ratio of your study region’s (a county geographic in the US) population to a weighted average of the four counties you purchased for Hawaii. After you compute this ratio, you would weight all of the impact results by this factor. The product would be an estimate for your region in your Malaysian region. Using population as the weighting factor would eliminate the need to have compatible geographic regions since population is a common denominator across all regions no matter the size. Another option would be to use Gross Regional Product (GRP) to weight all of your impact results. Since GRP reflects the size of a study region’s economy, weighting your region by the weighted average of the four counties in Hawaii would provide a way for you to compare the adjusted results using population and GRP as weights. From there, you could select the approach that seems to be most representative of your region in Malaysia. We hope this provide additional insight about how you might handle your situation.
    0
    Comment actions Permalink
  • Avatar
    Kei Wei
    I read the case study 12: Impacts of wildlife tourism. What's the different between commodity change and industry change? I had collected visitor's expenditure in the past few months. I am not sure whether I had classified the IMPLAN industry and commodities correctly (please see the attachment). The spending I obtained was based on 380 respondents while the total visitors for 2013 were 90,562. The case study 12 used US WI data so it will be no problem since all employment and industries data are from US and truly reflect the economic situations. But, I found a problem using the HI data (Maui, Kauai, Hawaii, and Honolulu). Which county economic is largely depend on tourism because my study area is a diving and honeymoon destination? However, the main income for the county is from agricultural. Let's say I use the Maui data, I see there is employment information, output and value added information, but some industries do totally not exist in my study area, e.g. mining, manufacturing. Also, I do not know the employment for each of the industries because the information is not available. If I generate the output, the employment information may be incorrect since the employment is from Kauai. What can I do to 'massage' the data to reflect my study region? Should I customize the industry production and commodity production? My study region has population 133,164, which is closer to the population of Maui 156,764. Therefore, I should choose Maui County. The average Household Income of Maui is USD100, 712. However, according to the latest statistics on 2012, the rural mean monthly gross household income was RM3080 or USD 963 (Malaysia) and RM4089 or USD 1278 (state of my study region). Also, I do not know the exact number of industries in my study region since it is a very small town. The size of my region is 442km2 but the economic activities I want to study is just about 10km2. The remaining 432km2 is palm oil estates or forest. If I use the population or GRP, what are the steps I should do? Again, I should stress this issue because my examiners committee is still not convinced that using IMPLAN US data outside the US. I have no choice since I had purchased the data. But I am sure I can do that by giving strong justifications. Although my study can be done using Input-output analysis, it doesn’t capture the full picture of the economy and IMPLAN is the extension of IO. Also, using IO analysis is very common and I would like to try something new using IMPLAN since it would definitely add value to my research. Please also refer to attachment for Malaysia IO tables. Looking forward to hearing from you soon.Thank you.
    0
    Comment actions Permalink
  • Avatar
    IMPLAN Forum
    Hi Kei Wei: We posted our comments in appropriate places within your recent Post. They are shown below. I read the case study 12: Impacts of wildlife tourism. What's the different between commodity change and industry change? Response: Impact on Commodity Sectors is divided on the basis of Market Share between all industries that contribute to a commodities production, the impact may be divided between several Sectors or may be "lost" from the model as Direct Institution Change if the Sector is produced in part by governments or other institutions. Unlike when you use a Commodity Change, with an Industry Change, there is no Direct Institution Change associated to your Events because you have indicated to the software that you are purchasing directly from Industries as opposed to purchasing the Commodity from all available producers. I had collected visitor's expenditure in the past few months. I am not sure whether I had classified the IMPLAN industry and commodities correctly (please see the attachment). The spending I obtained was based on 380 respondents while the total visitors for 2013 were 90,562. Recommendation: Total all spending by category for the 380 respondents. Then divide the spending in each category by the total number of persons associated with the 380 respondents (if expenditures were reported on a per party basis). That is, for each respondent there were an average of 3 persons in their party, then the total number of persons associated with these expenditures is 1,140. The reason why you want to convert your expenditures to per person expenses is because you have a total visitor count on a per person basis. So, if $15,000 were spent on overnight lodging (hotels and motels), then the average amount spent per person would be $36.29 person. However, if your expenditure data is on a per respondent basis, then you don’t need to worry about the earlier conversions. As far an attachment, we looked at those associated with your Post. First, we completed bridging your expenditures to the appropriate IMPLAN sectors for analysis purposes. I have attached it to the reply. We also looked at the I-O Matrix for Malaysia and the Gross Absorption Table as well. Reviewing such detail information and providing feedback goes beyond what we will do as a part of service and support through the FORUM. The case study 12 used US WI data so it will be no problem since all employment and industries data are from US and truly reflect the economic situations. But, I found a problem using the HI data (Maui, Kauai, Hawaii, and Honolulu). Which county economic is largely depend on tourism because my study area is a diving and honeymoon destination? However, the main income for the county is from agricultural. Let's say I use the Maui data, I see there is employment information, output and value added information, but some industries do totally not exist in my study area, e.g. mining, manufacturing. Also, I do not know the employment for each of the industries because the information is not available. If I generate the output, the employment information may be incorrect since the employment is from Kauai. What can I do to 'massage' the data to reflect my study region? Should I customize the industry production and commodity production? My study region has population 133,164, which is closer to the population of Maui 156,764. Therefore, I should choose Maui County. The average Household Income of Maui is USD100, 712. However, according to the latest statistics on 2012, the rural mean monthly gross household income was RM3080 or USD 963 (Malaysia) and RM4089 or USD 1278 (state of my study region). Also, I do not know the exact number of industries in my study region since it is a very small town. The size of my region is 442km2 but the economic activities I want to study is just about 10km2. The remaining 432km2 is palm oil estates or forest. If I use the population or GRP, what are the steps I should do? Again, I should stress this issue because my examiners committee is still not convinced that using IMPLAN US data outside the US. I have no choice since I had purchased the data. But I am sure I can do that by giving strong justifications. Although my study can be done using Input-output analysis, it doesn’t capture the full picture of the economy and IMPLAN is the extension of IO. Also, using IO analysis is very common and I would like to try something new using IMPLAN since it would definitely add value to my research. Recommendation: We would strongly recommend that you purchase the 2005 IMPLAN Data for Malaysia. This would at least give you a basis from which to make your assumptions and modifications for your study region. Although we offered you some guidance on how you might use the Hawaii counties to help in your research, we simply did not account for the fact that IMPLAN has data for Malaysia in 2005. Thus, we would advise against using a weighting approach as suggested in an earlier Forum Post and use the latest available data from IMPLAN Group, LLC or other foreign sources for your country. This would be a much better approach than using a weighting scheme from another country to proxy your study region and develop rough estimates of your results.
    0
    Comment actions Permalink
  • Avatar
    Kei Wei
    Thank you for your response. May I know the price of 2005 IMPLAN Data for Malaysia?
    0
    Comment actions Permalink
  • Avatar
    IMPLAN Forum
    Hi Kei Wei. As stated in our latest reply to you and your recent Forum Post, you may purchased the 2005 National Data package for Malaysia for $2,000. We can get you the data in one of two ways: 1)Via mail or 2) Schedule a remote LOGMEIN session where we could manually transfer the data over to your computer. We hope this answers your latest question. Please let us know if we can be of further help.
    0
    Comment actions Permalink
  • Avatar
    Kei Wei
    Thanks for your reply. Unfortunately $2000 is out of my budget. Since I had purchased HI data.Therefore, I think I have to use it for my analysis, otherwise it will be wasted.So, what should I do now?
    0
    Comment actions Permalink
  • Avatar
    IMPLAN Forum
    Hi Kei Wei. Given that you will not be able to purchase the 2005 data for Malaysia, about the only thing that we could suggest is that you follow some of the guidance that you have received from us in the last several Posts. You are in a difficult position with little to no data to model impacts of your projects in Malaysia. Whatever you do, we would suggests that you state your assumptions, what data you used from Malaysia and Hawaii, and what data you used from IMPLAN. From there, you should clearly lay out your methods and procedures in deriving your results. This will help others in understanding and reproducing your work in the future. We wish you the best in your research.
    0
    Comment actions Permalink

Please sign in to leave a comment.