Socioeconomic Impact Analysis

Back to List of Questions

12. How were the final impact estimates for income, job, and tax losses determined?

As noted in earlier FAQ responses, the initial regional level impact estimates for lost income, jobs, and taxes were possibly adjusted for the number of firms (i.e., within the mining and manufacturing water use categories), and were also adjusted for the degree of water shortage within the water use category. These calculations were made for each IMPLAN sector in the water use category within each county. Values/acre-foot were multiplied by the number of acre-feet short and summed across all IMPLAN sectors in the water use category for each county or WUG. This process implicitly takes into account the proportion of water used in each IMPLAN sector. The resulting impact estimates therefore reflected the variability of economic activity within each WUG. For example, a county with large mining water use for oil and gas exploration would have a higher impact estimate than one which focused on the less profitable sand and gravel quarrying, even if the two counties had the same degree of water shortage.

Back to List of Questions

13. Impact estimates for the irrigated agriculture sector relied upon IMPLAN data, yet employed a slightly different portion of that output to obtain the final impacts. What was the difference in approach, and what prompted that alteration in methodology?

IMPLAN output for value-added for the irrigated agriculture IMPLAN sectors included the value of both irrigated and dryland production within each county/region. Direct use of the IMPLAN estimates of value-added would therefore have overstated the value of water in the sector. An alternative approach was used where estimates of the total output in dollars (crop yield*price*acres) was calculated for the appropriate crops using data from non-IMPLAN sources. A second type of multiplier, available from IMPLAN, was then used to convert the total output ($) estimates by crop and county to the regional level estimates of income. Similar multipliers were also available to convert total output to the regional level estimates of jobs. Taxes were not considered in this water use category due to the presence of government subsidies. The resulting estimates of lost income and jobs per acre-foot were then used in the analysis. One additional alteration was employed. Variability in water use, weather, and economics in the irrigated agriculture category prompted use of the five year average total output/acre-foot of use to determine the initial value of water per acre-foot. Total output data for the years 2012-2016 was used in developing these values, and adjustments for the degree of water shortage and the aggregation techniques described in earlier FAQ responses were used in determining the final impact estimates for the irrigated agriculture water use category.

Back to List of Questions

14. An alternative income measure (non-IMPLAN based) was employed for the steam-electric power generation water use category. Why was the different methodology necessary, and how was the measure calculated?

The steam electric power generation category is very dissimilar to the other categories considered. In general, without enough cooling water, utilities would have to throttle back plant operations, possibly forcing them to buy costly power from other providers or to generate higher cost power at other plants under their control in order to meet customer demands. Direct use of the value-added estimates from IMPLAN was deemed less indicative of the damages incurred due to a drought, and the income measure used was the expected cost of power purchased using the day-ahead market price within Texas. Average day-ahead prices for the heavy drought year 2011 were used as proxy values for the cost of the purchases, and year 2016 average kWh/acre-foot of water estimates (by county) were used to determine the amount of electricity needed to be purchased for the acre-feet of needs in each county. The final cost for each need was determined as the product of the electrical price/kWh and the estimated kWh associated with the acre-feet of needs. No impact elasticity functions were employed for this water use category.

Back to List of Questions

15. What area of geographic influence is accounted for in the Impact Analysis, and are spillover impacts to other planning regions considered?

Water needs are estimated at the WUG-level, by region, for municipal water uses and at the county-level for the remaining water use categories. Regional level multipliers from IMPLAN were used to adjust the initial estimates of income, jobs, and taxes to reflect impacts on other sectors in the planning region only. Potential spillover effects to other planning regions do exist, but are not accounted for in the Impact Analysis. The Impact Analysis assumes that the drought occurs in Region X only. Backward linkages to input sectors in other regions are not accounted for. Similarly, forward linkages to relevant sectors which use Region X outputs are also not accounted for.

Back to List of Questions

16. What is consumer surplus, and how is it used in the Impact Analysis?

Consumer surplus is a welfare economics measure of well-being, and is a useful tool to assess the damages or benefits accruing to consumers when economic or resource availability conditions change. In the context of a drought, reduced water supplies adversely affect consumers as they cannot consume as much water as they desire, and lost consumer surplus is a measure of how much homeowners would be willing to spend to be able to consume water at those normal levels of use.

Impact analysis estimates of lost consumer surplus were made using household level, municipal WUG specific, water demand functions (both outdoor and indoor water use) based on average price and quantity of water used data from the Texas Municipal League annual survey data effort (year 2016). These estimates varied with the degree of shortage, up to shortages of 80%. Consumer surplus estimates for shortages greater than that value were not estimated, primarily due to estimation issues associated with the demand functions for very small quantities of water.

Back to List of Questions

17. Are impacts in the forward linked sectors for a drought impacted production sector also considered in the Impact Analysis?

Traditional use of input-output analysis via IMPLAN only captures impacts on backwardly linked sectors (i.e., wheat sales as an input to pasta production). Potential adverse impacts in the forward linked sectors, such as the reduced retail beef sales accompanying fewer marketable cattle following a drought, are not accounted for in the Impact Analysis. This is a shortcoming of the current approach, and is a facet which the IMPLAN model is not readily suited for, especially when dealing with numerous impacted sectors and 254 counties.

Back to List of Questions

18. Several factors caused differences in impact estimates when comparing results from the 2016 RWPs with the more recent 2021 RWPs. Updates of the relevant data, changes in impact assessment methodology, and the revision of needs all contributed to such impact estimate differences. How did those factors specifically contribute to changes in the lost income impact estimates when examining the state level impacts?

The table below depicts the general trends across decades for the magnitude and direction of changes for the income impact estimates when comparing the 2021 state totals vs the 2016 state totals. The magnitude of change is shown by the relative size of the arrows (small (<5%), medium (5 to 25%), and large (>25%)), and the direction and color of the arrow indicating whether a particular measure increased or decreased when comparing impact estimates for the 2021 RWPs versus similar impact estimates from the 2016 RWPs. Values shown are for the state as a whole using the sum of all the regional water plan estimates, with similar tables being possible for each planning region. Trend indicators are shown for one change factor (needs) and two possible analysis responses (lost income, and lost income/acre-foot of need).

Projected needs are smaller in five of the six categories shown, increasing only in the livestock sector. The accompanying income loss estimates increase in magnitude in four of the six water use categories, with small to medium declines in the overall projected income losses in the manufacturing and steam electric water use categories. The productive value of water, and the accompanying increase in adverse income losses, continues to increase over time as evidenced by the medium to large increases in the income losses/acre-foot of need shown in all six water use categories.

Note that the trends depicted in the figure above could likely vary significantly if portrayed for an individual planning region. Regional level trends may be masked when combined with other regional data and results shown at the state level. The magnitude, and possibly even direction, of a given trend indicator (arrow) might change due to differences in economic conditions or water use projections within the region of interest, for a given water use category, when compared to the state level results shown in the original figure.

Back to List of Questions

19. Why were projected demands in the manufacturing water use category markedly smaller in the 2021 RWPs versus the 2016 RWPs, and why, despite the decreased demands, were projected impacts larger in the 2021 RWP?

State totals reported for manufacturing water use have been on the decline in recent years, yet the value of output per acre-foot of water use has increased. Water use efficiency has increased. In the interest of providing conservative water use projections, year 2030 manufacturing water use projections were assumed to remain constant throughout the remainder of the planning horizon, despite recent downward trends in total use. The increases in water use efficiency, coupled with the increase in value of the water, outpaced the decreases in demands and subsequent needs, resulting in larger impacts in the 2021 RWP.

For more information regarding the projection methodology for manufacturing, please visit: Projections Methodology

Back to List of Questions

20. Why do some counties and water use categories have a water supply need, yet no economic impact?

Positive needs may exist for a given decade, yet use of the impact elasticity functions may result in no estimated economic damages. For example, if water shortages are less than 5%, the Impact Analysis assumes there is enough flexibility in the water supply system or production unit to result in no damage assessment.

Back to List of Questions

21. Why do impact estimates possibly differ for counties with similar volume of needs?

The variety of goods and services produced varies from one county to another, and the procedures used for determining the impact estimates reflect that variability as well as the differing value of water in various economic sectors and water use categories. For example, counties with large irrigated agricultural production will have smaller adverse impact estimates when compared to counties with a similar degree of need in manufacturing due to the higher value of water within manufacturing.

Back to List of Questions

22. How were utility revenue losses calculated?

Year 2016 data from the Texas Municipal League (TML) annual survey concerning average monthly water prices paid and quantity used was employed to estimate lost utility revenue for both residential and commercial use. Price data is available for both residential and commercial use, and average prices from the TML data were multiplied by the acre-feet of needs to estimate the value of lost utility sales. The TML data for various years, including the year 2016 values, is available at the TML Survey Results Website

Back to List of Questions

23. Projected water needs for the state under drought of record conditions more than double from 2020 to 2070, increasing from 3.1 million acre-feet to nearly 6.8 million acre-feet. The associated expected income losses (value added) increase by approximately 39% (from 98 to 136 billion dollars). Why the disparity in the magnitude of the impacts, with needs more than doubling but lost income impacts increasing by a smaller percentage amount?

Lost income effects vary markedly by water use category. When viewed at the state level with all six water use types combined, however, the presence of a large quantity of lower valued water (associated with nearly 3 million acre-feet of needs for the irrigated agriculture sector) dampens the overall possible lost income for the state. A doubling of needs by the year 2070, with a large proportion of lower valued water associated with irrigated agriculture, is not sufficient to double the lost income estimate from the initial year 2020 estimate for the state as a whole.

Individual factors which influence the final lost income impact estimates include the following:

1. Trends in water demands implicit within the demand projections: (i.e., fracking water use within mining is expected to decline in the later decades of the planning period, while increases in water use efficiency are expected to dampen demands in future decades for the manufacturing category). Both the timing and projected demand amounts influence the final income loss estimates. Additional detail on development of the water demand projections appear in the PDFs below:
   Population and Municipal Demand Projection Methodology
   Brief Summarized Methodologies for Non-Municipal Water Demand Projections
2. Greater income losses due to increased degree of shortages: (i.e., lost income impact estimates (per acre-foot of need) increase as the degree of shortage increases within each WUG, resulting in higher overall lost income impacts as the degree of drought worsens). Impact elasticity functions are used to phase in larger lost income impacts as the degree of drought worsens. Additional details appear in the impact estimate methodology description: Detailed Methodology for the 2021 Socioeconomic Impact Analysis
3. The underlying value of water within each water use category, and the relative proportion of needs within each water use category: total lost income impacts for the state will vary with the composition of the impacts for each of the water use categories. For example, large population growth in the municipal sector, including the increase in needs for the water-intensive commercial subportion of municipal use, resulted in very large increases in lost income for that category, especially as the degree of shortage increases over time. As seen in the table below, the income losses for this subportion of municipal use grew from approximately 2% in 2020 to nearly 32% of projected income losses in 2070.

Additional detail concerning the trends in needs and lost income for each water use category (for the period 2020 to 2070) and the state as a whole appear within the following table.

Back to List of Questions

24. Why are the final impact estimates considered to be conservative?

Impact estimates within the analysis are likely conservative (too small) for several reasons, with a partial list appearing below:

1. The analysis focuses on those sectors with adequate water use data;
2. consideration of only 1 year of drought;
3. no consideration of impacts on the forwardly linked sectors (both within the region of interest as well as impacts on adjoining regions);
4. no consideration of backward linked impacts on other planning regions;
5. does not consider the possible impacts of building moratoriums for those WUGs with significant long term water shortages; and
6. a lack of considering the increasing scarcity of water throughout the region, and the resulting impact on the use of technology, on the true value of water.

Back to List of Questions

25. What outside and internal peer reviews aided in alterations for the methodologies and/or reporting format used in the 2021 RWPs?

Following completion of the 2017 State Water Plan, three major entities were consulted concerning potential improvement in the methodology and reports associated with the Impact Analysis. In addition, details of the draft methodology for the 2021 RWPs were sent to the RWPGs in 2018 for review. Results of those efforts, coupled with an internal review by TWDB personnel, appear below.

Back to List of Questions

26. Should lost consumer surplus estimates from the residential water use impact measures be added to other monetary impact measures to obtain a total monetary impact?

Lost consumer surplus is not an out-of-pocket cost or foregone income measure similar to the lost income estimated for the remaining water use categories. Although lost consumer surplus values are estimated monetary losses, they are a different type of monetary impact, and ideally should not be added to other monetary estimates of loss.

Back to List of Questions

27. How were water hauling costs estimated?

Sample potable water hauling costs (trip lengths of 30 to 60 miles one way) from several locations within the state provided a baseline value of $35,000/acre-foot. These costs could vary significantly depending on the length of trip and the source of the trucked in water. All shortages exceeding the 80% acre-foot level of the projected demand were assumed to be trucked to the WUG to meet minimum sanitary and consumptive needs.

Back to List of Questions

28. Why was the degree of shortage of 80% selected as the tipping point for impact measurement to switch from lost consumer surplus to trucking costs?

This threshold was chosen since it approximates general minimum daily use values required to meet daily human consumption and sanitary needs. Water for shortages exceeding the 80% threshold is assumed to have to be trucked in at significant cost. The specific minimum use percentage for human basic needs will vary by city, depending on how efficient the populace is at conserving water (i.e., how low the baseline average GPCD (gallons/person/day) values are for the city). A second factor prompting the switch was the reduced reliability of the underlying indoor demand functions when estimating lost consumer surplus for large degrees of shortage. The demand functions employed become very steep (primarily due to the underlying mathematical functional form used) as one approaches small levels of water use, and experience has shown that estimates of lost consumer surplus are not reliable in such instances.

Back to List of Questions

29. Should regional level monetary impacts be summed across all regions to determine a state level impact estimate?

Aggregation of the results across some or all planning regions will result in an underestimate of the impacts, primarily due to regional level IMPLAN models not taking into account spillover impacts to counties and regions not included in the parent impact model for the region. In practice, however, results are often summed. Users of the results should take note of the potential underestimation concern, and inform their intended audience of that point. The degree of underestimation is not easily quantified due to the many complex points of interaction among the multiple flows of inputs and outputs across planning regions.

Back to List of Questions

30. Should impact estimates for regional total lost income for all WUGs be used as an estimate of the potential benefits of a water management strategy which mitigates all or a portion of an anticipated need?

The temptation in this setting is to divide the estimate of regional total lost income by the acre-feet of needs used to derive that lost income estimate, and then multiply by the acre-feet of needs mitigated by the proposed water management strategy. Regional total lost income estimates should not be used in this context as a means of determining the possible benefits of implementing a particular water management strategy. Several major reasons apply:

1. The regional total lost income estimates will include a composite impact, combining the various degrees of shortages and values of water for each of the WUGS and water use categories used to determine the regional total lost income estimate,
2. The underlying needs (percent shortage) used in estimating the RWP impact estimate will likely not correspond to the degree of shortage mitigated by the proposed water management strategy. The impact estimates vary with the degree of shortage of individual WUG and with water use category,
3. The proposed water management strategy may not address the needs in a single water use category, complicating assignment of the possible benefits to each category.

Individual water management strategies would likely offset needs in several WUGS and within several water use categories. Matching the specific needs of mitigation to the appropriate WUG and type of need, and then to the appropriate impact values, would likely be a daunting task.

Back to List of Questions

Navigate to the TWDB Planning Data Homepage