Rural Electrification Administration (United States, 1935 to 1994)

When private utilities would not string lines to thinly settled farms, Washington created the Rural Electrification Administration in 1935 and lent money to cooperatives and public bodies; nearly all farms got power and almost every loan was repaid, but co-op voting, rate-setting and board control later became a problem, especially for Black members in the South.

The record

What we could and could not open

We opened: the 1936 statute record at govinfo (49 Stat. 1363); the National Archives guide to REA records (Record Group 221); the EH.Net encyclopedia entry; an Asterisk magazine article on co-ops; a Richmond Fed Econ Focus article; the NBER abstract of Kitchens and Fishback (2013 working paper, 2015 article); Living New Deal's REA entry; a Southern Oral History Program page; Leifermann and Wehner, "A Question of Power" (Southern Changes, 1996); the Senate Agriculture Committee's current compilation of the Act (amended through 2019).

We could not open: Executive Order 7037 itself (the National Archives page did not mention it); the Senate and House votes on the 1936 Act; the Kitchens and Fishback full paper (abstract only); Lee, Miguel and Wolfram's 2020 study (named in the Richmond Fed piece); the 1949 telephone amendment text; any USDA civil rights complaint record. Claims from these are marked.

The problem

• In 1935, "fewer than 11 of every 100 U.S. farms were receiving central station electric service" (Living New Deal). The Richmond Fed says nearly nine in ten urban and nonfarm rural homes had power in 1930, against about one in ten farms.
• Investor-owned utilities were unwilling to serve sparse areas because expected profits were too low (EH.Net). Asterisk says utilities wanted farmers to pay the full upfront cost of lines and put construction costs at $1,356 per mile or more; the Richmond Fed says up to $2,000; Living New Deal says $1,500 to $2,000. By 1939 REA-financed lines were built for under $825 per mile (Richmond Fed; Living New Deal). The cost estimates are secondary and come from the utilities' side and the REA's side; we did not find an audit.

The agency and the statute

• EH.Net and Living New Deal both date the agency to Executive Order 7037, May 11, 1935. Living New Deal says it was created under the Emergency Relief Appropriation Act of 1935 and was first expected to give relief jobs and make grants with private companies; when utilities hung back it became a lender. The National Archives lists administrators Morris L. Cooke (1935 to 1937), John M. Carmody (1937 to 1939), Harry Slattery (1939 to 1945) and Claude R. Wickard (1945 to 1953), and says the agency ran 1935 to 1994 and that the Rural Utilities Service took over its loan programs.
• The Rural Electrification Act, approved May 20, 1936 (Living New Deal; the EH.Net entry calls it the Norris-Rayburn Act, for Sen. George Norris and Rep. Sam Rayburn). The statute record at govinfo (primary, opened) shows: Section 2 lets the Administrator make rural electrification loans and conduct studies. Section 3(a) lets the Reconstruction Finance Corporation lend the Administrator up to $50,000,000 for the year ending June 30, 1937; Section 3(b) authorizes $40,000,000 a year for each of the following nine years. Half the annual funds are allotted by each state's share of farms without central-station service, the other half unallotted with no more than 10 percent to any one state. Loans for generating plants and lines must be "fully amortized over a period not to exceed twenty-five years." Wiring and appliance loans are capped at five years. The Administrator must give preference to states, municipalities, "peoples utility districts," and cooperative, nonprofit or limited-dividend associations.
• The number $410 million that appears in the EH.Net entry and in Asterisk's description of the executive order does not appear in the 1936 statute text we read (about $50 million plus $40 million a year for nine years comes to $410 million, which may be the source; we did not verify). The Richmond Fed says Congress "initially set aside $100 million," which also does not match. We rely on the statute.
• Interest and terms differ by source. The statute: 3 percent on RFC loans to the Administrator; Section 4 and 5 loans at the average federal long-term borrowing rate of the preceding fiscal year. EH.Net says the rate matched the federal rate and was fixed at 2 percent from 1944, with 25-year terms. Asterisk says "20-year loans at 2.88% interest." The Richmond Fed says co-ops borrowed at 2 to 3 percent. We could not reconcile the 20-year figure with the statute's 25 years and do not use it.
• The 1936 Act allowed loans for home wiring and appliances (Living New Deal; Richmond Fed). The Richmond Fed cites Lee, Miguel and Wolfram (2020) for the finding that this credit was crucial; we did not open that study.
• A 1949 amendment authorized low-interest loans for rural telephone cooperatives (EH.Net; we did not open the amendment).

Results

• Progress: Asterisk says about 90 percent of farmers lacked electricity in 1935, nearly half had it by 1945, and all but 4 percent were connected by 1956. Living New Deal says rural electrification rose from about 11 percent to almost 97 percent by 1960. EH.Net says nearly all farms had service by the mid-1950s. By 1943 the REA had lent $466 million, financed 380,000 miles of line and served over a million consumers (Living New Deal). EH.Net says by end of 1938 "350 cooperative projects in 45 states supplied electricity to 1.5 million farms." That 1938 figure is higher than Living New Deal's 1943 count of "over a million consumers," so one of them is probably wrong. We do not use the 1938 figure.
• Repayment: EH.Net says the default rate was "less than 1%"; the Richmond Fed says "almost all REA loans were repaid, with a default rate below 1 percent"; Asterisk says almost every co-op repaid in full and on time. Kitchens and Fishback (NBER abstract) say the up-front interest subsidy was large but nearly all loans were repaid, so the taxpayer cost was "relatively low."
• Economic effects: Kitchens and Fishback (1935 to 1940) find REA loans "contributed significantly to increases in crop output and crop productivity," and that the number of electrified rural farm homes doubled within five years. Effects on nonagricultural activity were "much smaller."
• Opposition: Asterisk reports private utilities fought co-ops with "spite lines" meant to block co-op territory, high proposed taxes and propaganda, and that local officials and some state courts obstructed service. It also notes the REA-funded film Power and the Land was advertised as a documentary but was a scripted drama. This is one secondary account; we did not open primary records of spite lines.

Who was included and who was left out

• Structure. Co-ops were nonprofit, owned by customers, with boards elected from residential members. They could not serve people living in utility-served areas or in municipalities of 1,500 or more (EH.Net). Asterisk says the average co-op had about 800 members and 300 square miles.
• Race and board control. The Southern Oral History Program page says co-op boards "frequently shut out Black farmers and some white residents in remote areas" and calls service "uneven by both race and class." Leifermann and Wehner (1996) say each customer had a vote but white boards dominated the Black Belt through intimidation, misinformation and manipulated elections. Their examples: Delta EPA in Greenwood, Mississippi, 1983, where the all-white board walked out and re-elected itself two months later; Twin Counties EPA in Hollandale, 1983, where the manager held 569 proxies (270 eligible) and incumbents won 1,909 to 189. Organizers targeted 21 Black Belt co-ops in 1983 and only Dixie EMC (Louisiana) produced a board election win. Using 1995 USDA RUS data on 78 Black Belt co-ops: Black customers made up 25.8 percent of members, Black attendance at annual meetings was 35.2 percent, and Black board members were 9.1 percent. In January 1996 RUS formed a Cooperative Governance Team to address procedures that "appear to have been compromised"; the authors note it was not an enforcement unit. These are an advocacy-leaning regional magazine's numbers; we did not see the RUS data.
• Statute and discrimination. The current Senate compilation says the Secretary "shall give preference to" cooperatives, and its Section 9 bars political tests for staff. We found no nondiscrimination clause for borrowers in the text we read (the compilation, not the 1936 original), and we do not claim there was none in 1936.
• Tenants and renters. None of the sources we opened addresses tenant farmers. Unverified.

Who paid

Federal lending carried the program; the up-front subsidy was "large" and the later cost low because loans were repaid (Kitchens and Fishback abstract). EH.Net records a criticism that interest subsidies continued long after the Depression and that service would not fall if they ended. Members paid their own rates and wiring costs; Leifermann and Wehner say Delta residential users were reportedly charged more than twice what large farms paid for irrigation power. That rate claim is unverified beyond that article.

Timeline

• 1930 to 1935: About one in ten farms electrified (Richmond Fed; Living New Deal says fewer than 11 of 100 in 1935).
• May 11, 1935: Executive Order 7037 creates the REA.
• May 20, 1936: Rural Electrification Act approved.
• 1939: REA placed in the Department of Agriculture (Living New Deal); average line cost under $825 per mile.
• 1944: Interest fixed at 2 percent (EH.Net).
• 1949: Telephone loan authority (EH.Net).
• 1956: All but 4 percent of farms connected (Asterisk).
• 1983: Mississippi co-op elections described by Leifermann and Wehner.
• Jan. 1996: RUS Cooperative Governance Team.
• 1994: REA abolished; Rural Utilities Service takes over (National Archives). Living New Deal gives Oct. 13, 1994.

What this case shows

1. Markets left a group out, and a public program was the fix. Private utilities did not serve farms because the return was low. The REA lent to people who would build and own the lines themselves. Rural residents had no way to vote a utility into serving them; they could organize a co-op and borrow.
2. The co-op gave each member a vote on a service the member needed. The design was one-member boards chosen by customers. For a basic service this is unusual: the buyer owns the seller. This is as close as the case gets to "ordinary people could vote on and fix" and it is why the program is a strong example.
3. The public could check the result. The loans were mostly repaid and the figures on repayment come from several independent sources, including an economics working paper. A program of this size left a record anyone could audit.
4. But the vote was not equal in the Black Belt. The same structure that gave members a vote let boards use proxies and uncontested elections. Leifermann and Wehner give specific numbers (a manager holding 569 proxies; 3.1 percent annual-meeting attendance across Black Belt co-ops). A program with a ballot can still fail people who cannot use it.
5. The federal lender knew late. The USDA's own RUS data in 1995 showed Black board membership of 9.1 percent against 25.8 percent of customers, and the response in 1996 was a governance team, not an enforcement unit. We did not find an earlier federal review.
6. The distinction between the farm and the town mattered. Co-ops could not serve places inside municipalities of 1,500 or more. Who got service was drawn by population lines, not need.
7. Benefits were large but concentrated in farming. Kitchens and Fishback say effects on crop output and productivity were substantial while nonfarm effects were much smaller. Those who gained most were farm households and operators.

The strongest objections, and our answers

1. Objection: This was a success; the race problem is a 1980s story. Mitigation: both parts are true. We state the success (near-total electrification, loans repaid) and separately state the 1983 and 1995 evidence from one source. We do not claim the 1936 to 1950 co-ops excluded Black members at that time; the Southern Oral History Program page says boards "frequently" shut out Black farmers, but we did not find a count for the early decades.
2. Objection: The Southern Changes article is advocacy. Mitigation: yes, and it is the only source we have for the specific election and proxy numbers. It uses RUS data for the 1995 percentages. We label its claims and list RUS 1995 data as the primary to check.
3. Objection: Subsidies cost taxpayers a lot. Mitigation: EH.Net records exactly this criticism. Kitchens and Fishback say the up-front subsidy was large but taxpayer cost relatively low since loans were repaid. We did not find a dollar total for lifetime subsidy.
4. Objection: Private utilities would have served farms eventually. Mitigation: the Richmond Fed says utilities proposed a plan that exceeded the budget and fell short of full coverage; Asterisk says utilities blocked co-ops with spite lines. Neither source shows what unaided private build-out would have achieved, so the question is unresolved.
5. Objection: The loan terms numbers are inconsistent, so the entry is unreliable. Mitigation: correct that secondary sources disagree on interest (2 percent, 2.88 percent, 2 to 3 percent) and term (20 or 25 years). The statute says 25 years and sets rates by formula; we use it and leave the others flagged.
6. Objection: The REA propaganda film shows it manipulated the public. Mitigation: one secondary source says so. Both private utilities and the REA used propaganda by that account. We make no claim beyond that.
7. Objection: Co-ops are voluntary private bodies, not a federal failure. Mitigation: the REA lent public money to them and the later federal response (RUS governance team) shows the agency treated the problem as its concern.

Open questions: what would change our minds

• Executive Order 7037 and the 1936 Act's original text beyond the fiscal sections. Read the full Statutes at Large chapter, including any nondiscrimination or membership language.
• Funding figures. Reconcile $410 million (EH.Net, Asterisk), $100 million (Richmond Fed) and the statute's $50 million plus $40 million per year.
• Interest and term. Find REA annual reports for actual rates and terms by year.
• 1938 count. EH.Net says 1.5 million farms by end of 1938; Living New Deal says over a million consumers by 1943. Find the REA annual report.
• Farm percentages by year. Use Census of Agriculture; we have three secondary figures only.
• Race and tenancy in the early decades. Find records on who was served in 1936 to 1950, whether tenants and Black farmers were members, and whether early co-ops denied membership. Not verified.
• RUS 1995 data and the 1996 governance team. Primary USDA documents not opened.
• Kitchens and Fishback and Lee, Miguel and Wolfram. Read full papers; we read only an abstract and a secondary summary.
• Spite lines and the REA film. One secondary source.
• 1949 telephone amendment and 1994 abolition. Statutes not opened.
• If early records showed Black farmers were widely members and board participants in the 1940s, the Black Belt argument would change in scope; if tenants were largely excluded, we would add that.

Sources

Found an error or a better source? We keep every correction on the record with its date and reason. Nothing here is quietly rewritten.