Hospital union election activity, 1974-85

Abstract

This study, using National Labor Relations Board data and American Hospital Association data, reports on the status of union election activity in the hospital industry for a 65-month period, January 1980-May 1985, and contrasts it with earlier data for a similar 65-month time period (1974-79). Together these data provide a comprehensive overview of union election activity in non-Federal, nongovernment hospitals since the passage of the 1974 Nonprofit Hospital Amendments to the Taft-Hartley Act. The study analyzes union, election, hospital, and environmental characteristics. Comparisons over the two time periods show that, while union victory rates in hospital elections have remained constant, the total number of elections has declined dramatically in the hospital industry.

Introduction

With the enactment of the Nonprofit Hospital Amendments to the Taft-Hartley Act (THA) in August 1974, Congress brought the Nation's private (nongovernmental), nonprofit hospitals under the jurisdiction of the National Labor Relations Board (NLRB). This action affected more than 1.5 million hospital employees in more than half the Nation's 7,000 hospitals that were not previously covered under any Federal labor legislation. Several studies have explored the impact of union election activity in the hospital industry during the 1970's (Adamache and Sloan, 1982), but relatively little work has been done to provide an overview of union election activity in the 1980's. This study seeks to fill this gap by analyzing NLRB Monthly Election Reports for the health care industry for a 65-month period, January 1980 through May 1985. Combining this election information with data on hospital characteristics from the 1984 American Hospital Association (AHA) Annual Survey of Hospitals, this study gives a summary of union activity in the hospital industry for the first half of the 1980's. In addition, this investigation also combines the recent data with earlier work on union activity that appeared in the Health Care Financing Review and summarized union election activity from NLRB Monthly Election Reports and AHA data for the 65-month period August 1974 through December 1979 (Becker, Sloan, and Steinwald, 1982). Together these two studies give a comprehensive overview of union election activity in the hospital industry from 1974 to 1985.

The analysis begins with a brief background on the history of labor law and the hospital industry. Then the NLRB election results from 1974 through 1979 are summarized. Election information for hospitals and area characteristics for the 1980-85 period are analyzed and compared with the 1974-79 period. Union and election characteristics are similarly compared, and union elections and victories are examined by period and year. Finally, the summary and conclusions are presented.

Background

The Wagner Act of 1935, also known as the National Labor Relations Act (NLRA) implemented landmark Federal legislation that protected worker's right to organize and collectively negotiate with employers. It is the major Federal statute governing labor relations in the United States. The NLRA outlined employer's responsibilities and rights in the bargaining process. For example, the NLRA identified unfair labor practices (i.e. interference, coercion, descrimination against union members, not bargaining in good faith) and specified the general nature of how the employers should interact with union representatives once the union was recognized. In addition, the NLRA established the National Labor Relations Board (NLRB) and provided the NLRB with broad powers to oversee the Nation's union election activities and resolve unfair labor practices. The NLRA initially included all private hospitals, both nonprofit and profit. Government hospitals, however, were excluded.

In effect, many have argued that passage of the NLRA signaled a fundamental change in public policy and, consequently, power relations in the workplace. Furthermore, the NLRA gave government, as a regulator, a direct role in labor relations (Begin and Beal, 1985).

In 1947, the NLRA was amended by the Taft-Hartley Act (THA) to redress imbalances in the original legislation. While reaffirming the original intentions of the NLRA and the rights of employees to organize, THA identified unfair labor practices on the part of employees and expanded the role of the NLRB to cover these activities. THA also elaborated on the nature and composition of the bargaining unit. For example, some of the provisions in THA specify that supervisors cannot be part of the bargaining unit, security guards must be in separate bargaining units, professional and nonprofessional workers should generally not be combined in the same bargaining unit, and a 60-day notice of the contract termination or modification must be given by the parties involved.¹

One provision in the THA (Section 2.2), however, specifically excluded from its definition of employer “…. any corporation or association operating a hospital if no part of the net earnings inures to the benefit of any private shareholder or individual …..” (Rakich, Longest, and Darr, 1985). Also excluded in this section of THA were any government-owned corporations. As a result, most of the Nation's hospitals were now excluded from coverage by the NLRA, and only proprietary hospitals remained covered by THA. Nevertheless, at the NLRB's discretion, jurisdiction was not extended to proprietary hospital employees under THA until 1967 (National Labor Relations Board, 1978).

NLRA coverage was modified in 1962 when President Kennedy signed Executive Order 10988 authorizing collective bargaining in the Federal services. Federal hospitals were now afforded protection by labor law regulation. In return for the right to organize, Federal hospital bargaining units established under E.O. 10988 must agree not to strike. However, nonprofit and government hospitals were still not covered by any Federal legislation. With the enactment in August 1974 of the Nonprofit Hospital Amendments to the Taft-Hartley Act, Congress finally brought the Nation's private (nongovernmental), nonprofit hospitals back under the jurisdiction of the National Labor Relations Board.

When private, nonprofit hospitals were excluded from NLRA coverage in 1947, States were free to fill the legislative vacuum. However, most States opted not to take a position. By 1974, only 12 States had enacted laws to regulate hospital union activity.² Moreover, the provisions and coverage for union activity in these States varied considerably. As a result, the vast majority of hospital health care employees were not covered by the NLRA until the 1974 legislation (U.S. Department of Labor, 1979).

Since these changes in 1974, health care analysts and policymakers have focused considerable attention on the health care industry in an effort to understand how union election activity will influence the delivery and costs of health care (Becker, Sloan, and Steinwald, 1982). Although evidence indicated that only 27.4 percent of all U.S. hospitals were unionized in 1982 (Becker, Sloan, and Steinwald, 1982), one study predicted that by 1990, 65 percent of the Nation's hospitals would have a signed union contract (Feldman, Lee, and Hoffbeck, 1980). If correct, this increase could have a dramatic impact on hospital costs that for fiscal year 1984 exceeded $160 billion (Levitt, et al., 1985) and are projected to exceed $213 billion by 1988 (Arnett, et al., 1986). Sloan and Steinwald (1980), for example, found that a hospital that acquired an active union (one willing to strike or engage in other job actions) reported short-run labor costs 12 to 15 percent higher than those hospitals without an active union. Given that, in general, labor costs constitute 55 percent of a hospital's total budget³, the consequences of increased unionization in the hospital industry could be significant. But how have the unions fared since the passage of the 1974 amendments?

Union election results: 1974-79

In an earlier study, Becker, Sloan, and Steinwald (1982) found that 16.2 percent of U.S. nongovernment hospitals had elections during the period August 1974 through December 1979. Of these elections, unions won 48.6 percent. Seventy-one percent of the 1,025 elections occurred in three of the nine U.S. Census Divisions, Middle Atlantic, East North Central, and Pacific, with these three census divisions representing 70.7 percent of the 498 union victories in the period.

Other findings from the earlier study can be summarized as follows:

• For-profit hospitals had the lowest percentage of union elections (8.6 percent) among the various forms of hospital ownership, but exhibited the highest union victory rate (57.0 percent).

• The percentage of hospitals with an election increased with hospital bed size. Elections increased from a low of 7.8 percent for hospitals with a bed size of less than 100 beds to a high of 33.0 percent, for hospitals with more than 400 beds.

• Unions had lower victory rates (32.0 percent) in right-to-work States⁴ and higher victory rates (56.3 percent) in States where employees had been offered State protection prior to the 1974 changes to Taft-Hartley.

• Of the various employee organizations, independent unions and the Service Employees International Union were involved in the highest percentage of union elections (52.0 percent) between 1974 and 1979 with the independent unions winning the largest percentage of elections, 61.7 percent.

• Professional and/or technical unions were involved in the largest number of elections, 43.9 percent, and had a win rate of 57.1 percent. Departmental unions had the highest win rate (69.2 percent), and the combined professional and office unions had the lowest (33.3 percent).

• Finally, there was a short-lived spurt in hospital union election activity immediately after the 1974 amendments were enacted. The number of elections peaked in 1977 at 237 elections and fell to 145 elections in 1979. In absolute numbers, union victories peaked with 113 victories in 1975, or 22.7 percent of all the victories for the 1974-79 period. The number of victories fell to 70 in 1979, or 14.1 percent of the victories for the 1974-79 period.

Union election results: 1974-85

Hospital and area characteristics

Union election results for the two 65-month time periods, August 1974 through December 1979 and January 1980 through May 1985, are shown in Tables 1-3. Data are presented on elections and election outcomes and include only nongovernmental hospitals, because government hospitals are not covered by the NLRA. Observational units are the hospital in Table 1, the election in Table 2, and elections by year in Table 3. A total of 1,025 elections in 556 hospitals were reported for the 65-month period 1974-79 and 834 elections in 537 hospitals were reported for the 65-month period 1980-85.

Table 1. National Labor Relations Board elections and outcomes in nongovernmental hospitals, by selected hospital and area characteristics: August 1974-December 1979 and January 1980-May 1985¹.

Selected characteristic	Number of hospitals with elections		Percent of hospitals with elections2		Number of elections		Number of union victories		Union victories as a percent of elections
	1974-79	1980-85	1974-79	1980-85	1974-79	1980-85	1974-79	1980-85	1974-79	1980-85
All hospitals	556	537	16.2	12.8	1,025	834	498	397	48.6	47.6
Census division
New England	58	59	24.9	25.2	106	92	61	45	57.6	48.9
Middle Atlantic	156	172	31.4	30.5	310	303	163	149	52.6	49.2
South Atlantic	33	32	7.5	5.1	55	50	23	19	41.8	38.0
East North Central	121	93	20.1	13.0	221	131	91	51	41.2	38.9
East South Central	17	11	8.2	3.7	23	21	13	9	56.5	42.9
West North Central	27	26	6.7	5.4	51	33	25	16	49.0	48.5
West South Central	13	4	3.4	0.1	18	6	4	3	22.2	50.0
Mountain	25	22	12.5	9.2	39	32	20	14	51.3	43.8
Pacific	106	109	23.5	20.3	202	152	98	80	48.5	52.6
Puerto Rico3	—	9	—	20.0	—	14	—	11	—	78.6
Ownership
Nonprofit-religious	115	83	17.7	10.6	176	135	64	63	36.4	46.6
Nonprofit-non-religious	395	403	17.6	15.3	756	623	381	311	50.4	49.9
For-profit	46	51	8.6	6.4	93	76	53	23	57.0	30.3
Bed size
Less than 100	118	112	7.8	7.7	204	159	108	84	52.9	52.8
100-249	207	202	18.5	16.8	387	333	186	159	48.1	47.7
250-399	121	117	25.8	23.0	198	181	86	76	43.4	42.0
More than 400	110	106	33.0	20.4	236	161	118	78	50.0	48.4
Right-to-work
No	509	504	21.4	19.0	950	785	474	378	49.9	48.2
Yes	47	33	4.6	2.2	75	49	24	19	32.0	38.8
Worker protection
No	268	227	11.2	7.5	453	340	176	137	38.9	40.3
Yes	288	310	27.9	26.5	572	494	322	260	56.3	52.6

¹Data for the period 1974-79 appear in Becker, et al. (1982), Table 4.

²Based on census of AHA registered hospitals 1974 and 1984, respectively.

³Due to research design, Puerto Rico was not included in the 1974-79 study.

SOURCE: National Labor Relations Board: Monthly Election Reports for two 65-month periods, August 1974-December 1979 and January 1980-May 1985.

Table 3. National Labor Relations Board elections and outcomes in nongovernmental hospitals by year, August 1974 through May 1985¹.

Year	Number of union elections	Union elections as a Percent of all union elections		Number of union victories	Union victories as a percent of all union victories		Union victories as a percent of union elections
	By year	By year	By year for all years 1979-85	By year	By year	By year for all years 1974-85
Total August 1974-May 1985	1,859		100.0	Σ895		100.0	(895 ÷ 1,859 = 48.1)
August 1974-December 1979
Period total	1,025	99.9	55.1	498	100.1	55.6	48.6
19742	74	7.2	4.0	50	10.0	5.6	67.6
1975	236	23.0	12.7	113	22.7	12.6	47.9
1976	181	17.7	9.7	87	17.5	9.7	48.1
1977	237	23.1	12.7	109	21.9	12.2	46.0
1978	152	14.8	8.2	69	13.9	7.7	45.4
1979	145	14.1	7.8	70	14.1	7.8	48.3
January 1980-May 1985
Period total	834	99.9	44.9	397	100.0	44.4	47.6
1980	252	30.2	13.6	110	27.7	12.3	43.7
1981	214	25.6	11.5	100	25.2	11.2	46.7
1982	150	18.0	8.1	68	17.1	7.6	45.3
1983	87	10.4	4.7	48	12.1	5.4	55.2
1984	99	11.9	5.3	53	13.4	5.9	53.5
19853	32	3.8	1.7	18	4.5	2.0	56.2

¹Data for the period August 1974-December 1979 appear in Becker, et al. (1982), Table 5.

²August-December only.

³January-May only.

SOURCE: National Labor Relations Board: Monthly Election Reports for two 65-month periods, August 1974-December 1979 and January 1980-May 1985.

Table 2. National Labor Relations Board elections and outcomes in nongovernmental hospitals by selected union and election characteristics: August 1974-December 1979 and January 1980-May 1985¹.

Selected characteristic	Number of union elections				Percent of union elections2				Number of union victories				Union victories as a percentage of elections
	1974-79	1980-85			1974-79	1980-85			1974-79	1980-85			1974-79	1980-85
All hospitals	1,025	834			100.0	100.0			498	397			48.6	47.6
Employee organizations3
Independent Union	4313	175		287	430.5	18.5		30.4	4193	69		128	461.7	39.4		544.6
American Nurses' Association		112				11.9				59				52.7
Service Employees	220	159			21.5	16.8			103	73			46.8	45.9
District 1199	128	139			12.5	14.7			64	71			50.0	51.1
Teamsters	68	62			6.6	6.6			21	19			30.9	30.6
Operating Engineers	51	22			5.0	2.3			24	12			47.1	54.5
State, County and Municipal Employees6	45	26			4.4	2.8			17	5			37.8	19.2
United Food and Community Workers	40	59			3.9	6.3			13	24			32.5	40.7
Laborers International	29	13			2.8	1.4			7	9			24.1	69.2
Communication Workers	14	10			1.4	1.1			0	1			0.0	10.0
Office Employees International	13	17			1.3	1.8			6	7			47.2	41.1
Guard Workers	10	19			1.0	2.0			6	6			60.0	31.6
Others	94	131			9.2	13.9			44	42			46.8	32.1
Type of union
Industrial	227	161			22.1	19.3			66	59			29.1	36.6
Departmental	39	52			3.8	6.2			27	29			69.2	55.8
Craft7	—	31			—	3.7			—	19			—	61.3
Guard	35	31			3.4	3.7			22	16			62.9	51.6
Professional and/or technical	450	372			43.9	44.6			257	199			57.1	53.5
Office, clerical and other white collar	131	70			12.8	8.4			67	35			51.2	50.0
Combined professional and office	18	10			1.8	1.2			6	3			33.3	30.0
All others	125	107			12.2	12.8			53	37			42.4	34.6
Type of election
Stipulation	597	575			58.2	68.9			293	281			49.1	48.9
Regional director-ordered	314	226			30.6	27.1			133	97			42.4	42.9
Board ordered	44	8			4.3	1.0			19	2			43.2	25.0
Consent	70	20			6.8	2.4			53	15			75.7	75.0
Expedited8	—	5			—	0.6			—	2			—	40.0
Nature of election
Single union	886	728			86.4	87.3			419	334			47.3	45.9
Multi union	139	106			13.6	12.7			79	63			56.8	59.4

¹Data for the period 1974-79 appear in Becker, et al. (1982), Table 5.

²Percentages do not always add to 100.0 due to rounding.

³Total “employee organizations” for the 1980-85 period totals 944 due to multiple employee organizations seeking to represent a category of employees at the time of a single given recognition election. Data for the 1974-79 period does not use this counting method.

⁴ANA data separate from the independent union category were not available in the study covering the years 1974-79.

⁵(69 + 59) ÷ (175 + 112) = 44.6 percent.

⁶Listed as Retail Clerks International Association in the 1974-79 study. In 1979, it merged with the Meat Cutters and Butchers Union to form the United Food and Commercial Workers' Union.

⁷Due to research design, craft type of union was not used as a separate category in the 1974-79 study.

⁸Due to research design, expedited elections were not used as a separate category in the 1974-79 study.

SOURCE: National Labor Relations Board: Monthly Election Reports for two 65-month periods, August 1974-December 1979 and January 1980-May 1985.

As shown in Table 1, 12.8 percent of nongovernment hospitals had an election during the January 1980 through May 1985 period. This is down slightly from the 16.2 percent during the 1974-79 period. Of the 834 elections, unions won 47.6 percent. This is almost identical to the earlier rate of 48.6 percent for the 1974-79 period.

The census divisions with the highest number of elections during the 1980-85 period were, in rank order, the Middle Atlantic (303), Pacific (152), East North Central (131), and New England (92) census divisions. These divisions accounted for 81.3 percent of all elections during the period and 81.9 percent of all union victories. When compared with the 1974-79 period, the percent of victories to elections fell in all census areas except the Pacific and the West South Central census division, where it rose from 48.5 to 52.6 and 22.2 to 50.0 percent, respectively. It should be noted that the sizable increase in the West South Central census division represents victories in only three of six hospital elections.

For the 1980-85 period, nonprofit, nonreligious hospitals were the most likely to experience an election (15.3 percent) and also the type of hospital ownership where the unions were most likely to win (49.9 percent). In both periods studied, for-profit hospitals still were the least likely among the various forms of hospital ownership to experience a union election (8.6 percent in 1974-79 and 6.4 percent in 1980-85). However, in contrast to the earlier evidence from the 1974-79 period that showed union success to be highest in for-profit hospitals (57.0 percent), it fell appreciably in the 1980-85 period to 30.3 percent, the lowest among the various forms of hospital ownership.

The number of hospitals experiencing a union election in both time periods appears to grow with bed size. However, for the latter period, it peaked at the 250-399 bed size. The less-than-100 bed size category still appears to be the hospital size where unions are most likely to be victorious (52.8 percent). This victory rate has changed little since the 1974-79 period when it was 52.9 percent.

In the 1980-85 period, unions continued to avoid areas where the legal and social environment (primarily the south) were not receptive to union election activity. For example, only 33 of the 537 hospitals with elections, or 6.1 percent, were in States with right-to-work (RTW) laws. These 33 hospitals represented only 2.2 percent of the potential hospitals in these States that were eligible for elections. This is slightly lower than the 1974-79 period when 4.6 percent of the hospitals in RTW States had elections. In contrast, States in which the legal environment has traditionally been receptive to union election activity—those with laws that protected and facilitated collective bargaining in nonprofit hospitals before 1974—had election rates that were more than three times higher than States that had no such laws (26.5 percent versus 7.5 percent).

Union and election characteristics

Frequency distributions of NLRB union elections by union and election characteristics as well as union victory rates for both the 1974-79 and 1980-85 periods are shown in Table 2.

The four employee organizations that were involved in the largest share of elections in the 1980-85 period were the independent unions (employees not associated with any national union), Service Employees International Union, District 1199 of the Retail, Wholesale, and Department Store Union, and the American Nurses' Association. Together these four unions accounted for 62 percent of all hospital elections.

The independent unions had a much lower victory rate, 39.4 percent, than the national average, 47.6 percent. In part, this lower victory rate, when compared with the 1974-79 study victory rate of 61.7 percent, is attributed to the fact that we have separated the American Nurses' Association (ANA) out of this category. The NLRB did not distinguish elections that the ANA was involved in until 1977. Nevertheless, even if the ANA and independent union categories are combined as they were in the 1974-79 period, the independent unions (plus the ANA) had a much lower union victory rate (44.6 percent) than they did in the 1974-79 period.

Unions organized into professional and/or technical units were responsible for almost half the NLRB elections during the 1980-85 period (44.6 percent), virtually the same percentage they participated in during the 1974-79 period (43.9 percent). This was more than twice the election rate of industrial employee organizations (19.3 percent), the type of union with the second highest number of elections. Professional and/or technical units also appeared to be above the national average in their victory rates. For the 1980-85 period, professional/technical units were victorious in 53.5 percent of their elections, considerably higher than the 47.6 percent national average. This is down slightly, however, from their 1974-79 victory rate of 57.1 percent.

The NLRB makes distinctions among five types of representative elections. Most common are the two kinds of consent elections. In the “agreement for consent” election, the regional NLRB director makes the final resolution in any disputes concerning the conduct of the election. In the “stipulation for certification on consent” election, the NLRB settles all disputes. The term “stipulation” in Table 2, is far more common than the former, termed “consent” in the table. Stipulation elections usually require a longer period of time to be resolved if disputes arise. This is typically because consent elections are not contested by the organization being unionized and therefore disputes are not as acrimonious. The fact that the NLRB is involved may appeal to unions and employers alike on “fairness” grounds. Nevertheless, union victory rates are substantially higher in consent elections than in stipulation elections, although involvement of the NLRB is not necessarily a factor in this difference.

Union elections may also be ordered by the NLRB or by a regional NLRB director. A Board-ordered election occurs when there are questions concerning the appropriateness of the bargaining unit or circumstances involving a novel issue, for example, a unique bargaining unit. The regional director may order an election when a disagreement occurs between the bargaining parties, but there are no novel issues or unique circumstances involved.

The final type of election identified by the NLRB is the expedited election. This usually occurs when an unfair labor practice is involved. In these situations the NLRB typically changes the election timeframe by moving the election forward.

There were only five expedited elections and eight Board-ordered elections in the 1980-85 period. The majority of elections were stipulated, followed by regional director-ordered elections, and consent elections, respectively. Although consent elections represent only a small proportion of the elections held during this period (2.4 percent), unions won 75.0 percent of these elections. Stipulated elections (representing 68.9 percent of all elections) had the next highest union victory rate, 48.9 percent, followed by the regional director-ordered election (representing 27.1 percent of all elections) with a union victory rate of 42.9 percent.

The vast majority of elections involved only one union. Multiple elections, in which two or more unions competed with one another to represent the bargaining unit, occurred approximately one-seventh as often as single union elections. However, in spite of this difference, multiple union elections had a higher victory rate (59.4 percent) than their single union counterparts (45.9 percent). These victory rates for multiple and single union elections did not change substantially from the 1974-79 period when they were 56.8 percent and 47.3 percent, respectively.

Elections by period and year

Union elections and victory data by period and by year are presented in Table 3. The number of union elections in the 65-month 1974-79 period was approximately one-fourth higher than the 65-month 1980-85 period (1,025 and 834, respectively) as was the number of union victories (498 and 397, respectively). The overall victory rates for both periods, however, were nearly identical—48.6 and 47.6 percent, respectively.

When combining data from both periods and examining all years, several observations should be noted. The highest percentage of the 1,859 union elections in the combined 10-year and 10-month period occurred in 1980. For this year, union elections as a percentage of total elections reached a high of 13.6 percent. In 1975 and again in 1977, the percentage of elections was slightly lower, 12.7 percent. Since 1981, however, the percentage of union elections taking place has fallen dramatically. In the last 2 full years of data, 1983 and 1984, the percentages of the 1,859 elections that occurred were 4.7 and 5.3 percent, respectively.

A similar pattern appears for union victories. Of the total 895 union victories that occurred between 1974 and 1985, 12.6 percent were in 1975, 12.2 percent in 1977, and 12.3 percent in 1980. In contrast, of the total number of union victories, only 5.4 and 5.9 percent occurred in the years 1983 and 1984, respectively.

Summary and conclusions

This study, using data from the National Labor Relations Board Monthly Election Reports for a 65-month period, January 1980 through May 1985, and the 1984 American Hospital Association Survey of Hospitals, has sought to update the status of union election activity in nongovernmental hospitals in the first half of the 1980's and compare that activity to the previously reported 65-month 1974-79 period. As a result, a comprehensive profile of union election activity in the hospital industry following the passage of the 1974 amendments to the Taft-Hartley Act is presented. A number of conclusions seem warranted.

First, in contrast to earlier predictions that 65 percent of all hospitals would be unionized, our evidence shows a rather dramatic decrease in the extent of union election activity in the hospital industry, especially since 1981. Our data do not permit us to estimate exactly how many hospitals were unionized by 1985 because a hospital can have more than one union. However, if we assume that the ratio of hospital elections to total elections is roughly the same as it was for the 1974-79 period (a very liberal estimate, because the “easier” hospitals were supposedly unionized first), we would calculate that slightly more than half of the 537 hospitals in our study that had elections (268 hospitals) were hospitals that were experiencing their first election. Since unions won approximately half of the elections they were involved in, this would mean that 134 hospitals acquired a union. This would represent approximately a 3.3-percent increase in the number of hospitals with unions since 1980. Adding this 3.3 percent to the 27.4 percent of hospitals that were found to have union contracts in 1980 (Becker, Sloan, and Steinwald, 1982) would suggest that approximately 30.7 percent of the Nation's hospitals had a union by May 1985. This is considerably lower than previous estimates. Moreover, it would imply that, without significant changes in the rate of union election activity, unions will have contracts in only about 34 percent of U.S. nongovernmental hospitals by 1990.

Second, somewhat surprisingly, union victory rates in the hospital industry have not declined substantially. For the 1974-79 period, union victory rates were 48.6 percent compared with 47.6 percent for the more recent 1980-85 period. This fact coupled with the evidence that unions are not as active in the hospital industry as in the earlier period (834 versus 1,025 elections) suggests that unions are being very selective in the hospitals they choose to unionize. Consequently, they are still able to win approximately half the elections they enter. This probably reflects the reality that the “easier” hospitals have been unionized and those remaining require considerably more time and resources if the unions are to maintain their success.

Third, the dramatic decline in the number of elections since 1981 would suggest that the hospital industry, similar to other industries, has become more resistant to union election activity and better able to oppose union organizing efforts. It would appear that the dramatic changes in the health services industry beginning in the early 1980's have had a substantial impact on hospitals. Included in these changes are the implementation of a prospective payment system for hospitals⁵, the dismantling of national health planning and its regulatory influence, and the rising competitive environment which includes health maintenance organizations, preferred provider organizations, and hospital diversification and consolidation. The net effect is that hospitals are more concerned with their productivity and costs (i.e., survival). It would appear that this hostile and turbulent environment with accompanying declines in occupancy, length of stay, hospital beds, and staff has made hospitals even more resistant to union organizing efforts. The fact that only 20 of the 834 union elections held during the 1980-85 period were consent elections (elections where the hospital did not contest holding the election) reflects the combative nature of the vast majority of these elections. Evidence has shown that where the election process is prolonged, such as in a stipulated or Board-ordered election in comparison to a consent election, unions' chances for success decline substantially (Prosten, 1978).

Fourth, different hospital and area characteristics have a dramatic impact on both the likelihood of an election and the outcome of the election. In part, the regional patterns reflect the concentration of hospitals in certain areas of the country. However, considerable social influence is still apparent. For example, regions of the country that have a strong history of supporting union election activity, as evidenced by States that had legislation supporting union activity in hospitals prior to the passage of THA, are more likely to have an election and the union appears more likely to be victorious than areas where such support is lacking. The converse appears to be true in areas that traditionally have not supported union election activity. In States that have right-to-work legislation, for instance, there is a much lower likelihood of having a union election and less chance that the union will win the election. This is very similar to the pattern that appeared in the 1974-79 period.

Fifth, the majority of the union elections since 1980 took place in nonprofit, nonreligious hospitals. This same group of hospitals, when compared with nonprofit religious hospitals and for-profit hospitals, experienced the highest union victory rate of 49.9 percent. This victory rate is almost identical to the union victory rate in this form of hospital ownership during the 1974-79 period of 50.4 percent. In contrast, the most dramatic change among union victory rates in the various types of hospital ownership for the 1974-79 versus the 1980-85 period is in the for-profit hospital. During the earlier period, unions had a 57.0 percent victory rate in for-profit hospitals; in the latter period, the union victory rate in for-profit hospital ownership dropped to 30.3 percent. This may, in part, result from the acquisition of many of the freestanding for-profit hospitals by multihospital chains. The fact that most of these multihospital systems have experienced labor relation departments and a larger pool of resources available to resist organizing efforts may account for the decline in union success between the two periods.

Sixth, among the various employees' organizations, there is still considerable variation in their success rate. Independent unions, which were responsible for the largest number of elections of any group in the 1970's, still accounted for the largest number of elections in the 1980's. Their victory rate, however, has declined dramatically from 61.7 percent in the 1970's to 39.4 percent in the 1980's. This may be explained in part by the fact that the elections of the American Nurses' Association (ANA) were combined with the independent unions category in the 1974-79 period. Nevertheless, even if they are combined, the independent unions still achieved only a 44.6 percent victory rate, considerably below their rate of victory in the 1970's.

Finally, in 1983 and 1984, the last 2 full years for which we have information, the number of union elections dropped below 100 for the first time since 1974. For the first 5 months of 1985, only 32 hospital elections had taken place. If this rate continues, it would imply that the hospital industry would experience around 77 elections in calendar year 1985. These figures are far below the 252 and 214 hospital elections that took place in 1980 and 1981, respectively. It would appear, at least for the near future, that union election activity will continue to decline in the hospital industry and unions will continue to experience considerable difficulty in organizing the hospital industry.