Abstract
Water can be a scarce resource, particularly in certain places at certain times. Understanding both water use and conservation efforts can help ensure that limited supplies can meet the demands of a growing population and economy. This paper examines water use and recirculation in the U.S. manufacturing sector, using newly recovered microdata from the Survey of Water Use in Manufacturing, merged with establishment-level data from the Annual Survey of Manufactures and the Census of Manufactures. Results suggest that water use per unit of output is largest for larger establishments, in part because larger establishments use water for more purposes. Larger establishments are also found to recirculate water more — satisfying demand (water use) without necessarily increasing water intake. Various costs also appear to play a role in water recirculation. In particular, the water circulation rate is found to be higher when water is purchased from a utility. Relatively low (internal) prices for self-supplied water could suppress the incentive to invest in recirculation. Meanwhile, establishments with higher per-gallon intake treatment costs also recirculate more, as might be expected. The cost associated with water discharge – due to regulation or otherwise – also increases circulation rates. The aridity of a locale is found to have little effect on circulation rates.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
According to the U.S. Geological Survey (Maupin et al. 2014), in 2010, 268 million people were served publicly with 23,800 million gallons per day for “domestic use” and an additional 44.5 million people were self-supplied with 3600 million gallons per day. This implies a usage of 87.7 gal per day per person. Meanwhile, manufacturing industries self-supplied 15,900 million gallons per day in 2010. In 1995, about 82 % of industrial water was self-supplied (Solley et al. 1998), the last year this estimate was made. Assuming this same proportion, total manufacturing water intake in 2010 was approximately 19,390 million gallons per day.
Data on the delivery of water by such suppliers was last published for 1995 (see Solley et al. 1998). In that year, the U.S. manufacturing sector self-supplied 22.4 billion gallons of water per day, or about 5.6 % of total water withdrawals in the United States. Another 4.8 billion gallons, or 1.2 %, came from public and private water suppliers. In 1973 and 1978 – the years studied in this paper – manufacturers self-supplied 89 % and 82 % of their water, respectively.
Total water intake in 1973 and 1978 was 15,024.3 and 12,991.7 billion gallons, respectively, while gross water use was 43,412.6 and 44,494.1 billion gallons, respectively.
Water used for hydroelectric power generation is not withdrawn and is not included here.
Even in today’s dollars, this represents hundreds of gallons of water per dollar of output.
This publication cautions us that, because of certain reporting inconsistencies, the 1983 water use and reuse statistics “are of lower reliability than the water intake and water discharged statistics presented in this report.” (See the online appendix for further details.) It is not clear whether these same concerns should perhaps apply to data from other years as well. No similar concern was raised in those publications. As will be discussed in the next section, the questions on gross water used and water recirculated/reused took different forms between 1983, 1978, and 1973.
As discussed in the online appendix, there is puzzling ambiguity on this point. Nevertheless, here and throughout, I will assume the strict identity: gross water used is the sum of water intake plus water recirculated/reused.
This paper will not separately consider water consumed in process — either embedded in the final product or through evaporation. This can be derived by subtracting water discharged from water intake. According to U.S. Bureau of the Census (1986), between 6 to 11 % of the water intake [2.0 % and 3.7 % of gross water used] by the manufacturing sector was consumed, depending on the year. That 90 % or more of water intake is eventually discharged by manufacturers (usually into rivers, streams, and other surface waters, less typically into public utility sewers or to the ground) may suggest that consumption is of more relevance than water intake or gross water use, especially if water availability is of particular concern. However, water quality is also important, in which case the quantity of water intake, its function within the manufacturing plant, and its treatment before discharge are all important too. Unfortunately, the SWUM does not have ideal measures of the quality of discharged water. We do know that untreated water accounted for 87 %, 77 %, 71 %, 70 %, 56 %, 60 %, and 55 % of discharged water in 1954, 1959, 1964, 1968, 1973, 1978, and 1983, respectively (U.S. Bureau of the Census 1986). However, “untreated” does not necessarily imply net degradation in quality. Likewise, “treated” does not imply the opposite. In any event, reductions in water intake seem unambiguously good by leaving more water untouched. And one would expect reductions in water intake to come mostly from greater recirculation than from reductions in consumption. The study of residential water use also tends to focus on intake and use rather than consumption. Here, too, most water is discharged.
Meanwhile, on the 1983 survey form, water intake and water recirculated/reused was asked and gross water used is implicit. In the previous section, I noted a caution that appeared in the publication regarding the 1983 water use and reuse statistics. Perhaps the different form of this question (where gross water use was not explicitly asked) led to confusion.
Also recovered was data from 1973 and 1978 for about 1690 and 1060 establishments in mineral industries, respectively. These observations are not employed in this current study.
I do not present the results of this regression here. The coefficients from this regression are biased since the sample only contains establishments that have at least 20 million gallons of water intake. With the left part of the water use distribution missing, water use intensity will be overstated – particularly for the smallest establishments and, no doubt, attenuating as establishment size increases. Nevertheless, in spite of the bias, water use intensity is found to increase over the last three establishment size categories and the highest water use intensity is unambiguously (at least in 1973) among the very largest establishments (with 2500 or more employees).
In the 1973 sample, the percent in each establishment size category is 19 % (1–99 employees), 26 % (100–249 employees), 21 % (250–499 employees), 18 % (500–999 employees), 11 % (1000–2499 employees), and 5 % (2500+ employees). In the 1978 sample, the percentages are 15 %, 27 %, 24 %, 18 %, 11 %, and 5 %, respectively.
Following Halvorsen and Palmquist (1980), the marginal effect of a dummy variable when the dependent variable is measured in logs is calculated as exp(α) – 1.
As noted earlier, “untreated” does not necessarily imply net degradation in quality, and “treated” does not necessarily imply the opposite.
The Census Bureau defined 20 such regions. As discussed in the relevant publications: “The boundaries for ... industrial water-use regions are combinations of counties without regard for State lines. These regions, which center around river basins, were developed in consultation with the Federal agencies having responsibility for water development or use and with representatives of private industry and they approximate the major drainage systems in the country.”
North America Land Data Assimilation System (NLDAS) Daily Precipitation 1979–2011, as extracted from the CDC WONDER Online Database on August 27, 2013.
This log TFP is a beta version constructed and employed by Foster et al. (2016), using longitudinal establishment-level data from the ASM and CM, together with aggregate data from other sources, including the Bureau of Economic Analysis and the Bureau of Labor Statistics.
References
Becker RA, Grim C (2011) Newly recovered microdata on U.S. manufacturing plants from the 1950s and 1960s: some early glimpses. Center for Economic Studies Discussion Paper Series 11–29
Bhat VN (2015) Water and its effect on business productivity: a cross-country analysis. Water Resour Manag 29(11):4007–4020
Bruneau JF, Renzetti S (2010) Water use intensities and the composition of production in Canada. J Water Resour Plan Manag 136(1):72–79
Bruneau JF, Renzetti S (2014) A panel study of water recirculation in manufacturing plants. Can Water Resour J 39(4):384–394
Bruneau JF, Renzetti S, Villeneuve M (2010) Manufacturing firms’ demand for water recirculation. Can J Agr Econ 58(4):515–530
Dupont D, Renzetti S (1998) Water use in the Canadian food processing industry. Can J Agr Econ 46(1):83–92
Dupont DP, Renzetti S (2001) The role of water in manufacturing. Environ Resour Econ 18(4):411–432
Ercin AE, Aldaya MM, Hoekstra AY (2011) Corporate water footprint accounting and impact assessment: the case of the water footprint of a sugar-containing carbonated beverage. Water Resour Manag 25(2):721–741
Fishman C (2016) Water is broken, data can fix it. New York Times [editorial], March 17, 2016
Foster L, Grim C, Haltiwanger J (2016) Reallocation in the Great Recession: cleansing or not? J Labor Econ 34(S1):S293–S331
Freire-González J (2011) Assessing the macroeconomic impact of water supply restrictions through an input-output analysis. Water Resour Manag 25(9):2335–2347
Halvorsen R, Palmquist R (1980) The interpretation of dummy variables in semilogarithmic equations. Am Econ Rev 70(3):474–475
Ku SJ, Yoo SH (2012) Economic value of water in the Korean manufacturing industry. Water Resour Manag 26(1):81–88
Maupin MA, Kenny JF, Huston SS, Lovelace JK, Barber NL, Linsey KS (2014) Estimated use of water in the United States in 2010. U.S. Geological Survey Circular 1405
Padowski JC, Jawitz JW (2012) Water availability and vulnerability of 225 large cities in the United States. Water Resour Res 48(12):1–16
Renzetti S (1992) Estimating the structure of industrial water demands: the case of Canadian manufacturing. Land Econ 68(4):396–404
Renzetti S (1993) Examining the differences in self- and publicly supplied firms’ water demands. Land Econ 69(2):181–188
Renzetti S (2015) Economic analysis of industrial water use. In: Dinar A, Schwabe K (eds) Handbook of water economics, Edward Elgar Publishing
Solley WB, Pierce RR, Perlman HA (1998) Estimated use of water in the United States in 1995. U.S. Geological Survey Circular 1200
U.S. Bureau of the Census (1976) 1972 Census of Manufactures: subject and special statistics, volume 1
U.S. Bureau of the Census (1981) 1977 Census of Manufactures: subject statistics, volume 1
U.S. Bureau of the Census (1986) 1982 Census of Manufactures: water use in manufacturing, subject series, MC82-S-6
van Oel PR, Hoekstra AY (2012) Towards quantification of the water footprint of paper: a first estimate of its consumptive component. Water Resour Manag 26(3):733–749
Acknowledgments
This paper has benefited from helpful feedback on earlier versions, including comments from Lucia Foster, Joseph Shapiro, and session participants at the Association of Environmental and Resource Economists (AERE), International Water Resource Economics Consortium (IWREC), and Comparative Analysis of Enterprise Data (CAED) conferences. Any opinions and conclusions expressed herein are those of the author and do not necessarily reflect the views of the U.S. Census Bureau. All results have been reviewed to ensure that no confidential information is disclosed. The author declares that he has no conflict of interest.
Author information
Authors and Affiliations
Corresponding author
Electronic Supplementary Material
ESM 1
(PDF 931 kb)
Rights and permissions
About this article
Cite this article
Becker, R.A. Water Use and Conservation in Manufacturing: Evidence from U.S. Microdata. Water Resour Manage 30, 4185–4200 (2016). https://doi.org/10.1007/s11269-016-1414-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11269-016-1414-7