Organized Labor

Supervisor John Avalos’s proposed local hire legislation will mandate from the start that 30 percent of City-funded construction jobs go to City residents. In the second year, it mandates 40 percent; in the third, 50 percent. To my knowledge, no one on the Board of Supervisors, including and especially Supervisor Avalos, has examined what this means mathematically.

Here is some of the math. In its course, I use the same studies on which Supervisor Avalos has relied, the “Labor Market Analysis” of L. Luster and Associates (LMA report) and the Brightline Defense – Chinese for Affirmative Action report, “The Failure of Good Faith” (Brightline/CAA report).

Workforce demand will fluctuate at a much more rapid rate than resident journeylevel workforce availability. The LMA report, for example, shows a workforce projection on City-funded projects of 9,389 Full Time Equivalent (FTE) positions in 2011, but a drop to 5,600 in 2012, a climb to 7,013 in 2013, and a return to more than 9,000 by 2014. Beyond these annual fluctuations, there will be spikes and dips in demand on monthly, weekly, even daily bases, both broadly and within specific trades. Numbers of resident journeylevel workers, on the other hand, will depend on apprenticeship cycles that will be three to five years, depending on trade. Changes in resident journeylevel workforce are further constrained by regulatory limits on apprentice use, which I discuss below. Relative to short-term workforce demand, then, the number of resident journeylevel workers will vary so much more slowly that we can treat that number as a constant.

Let j stand for number of resident journeylevel workers. This can be expressed as a percentage of total workforce demand. Let x stand for a decimal factor representing percentage, and w for short-term workforce demand in number of journeylevel workers. Then

j = xw
And so of course
j/w = x

The percentage is therefore in inverse proportion to short-term workforce demand, and percentage requirements will be more difficult to meet as short-term demand increases. This represents a basic conceptual flaw in the Avalos legislation.

Some argue that the answer to this is apprenticeship. Apprentice hours on public works are in fixed ratio to journeylevel hours, by state labor code; for most trades, this ratio is 1:5. Numbers of apprentices versus numbers of journeylevel workers will approximate the specified ratio. Let r stand for a fixed ratio, a stand for number of apprentices, and t stand for non-resident journeylevel workers, or travelers. Total short-term demand in numbers of workers can be expressed as resident journeylevel workers plus non-resident journeylevel workers, so:

j + t = w
Apprentice ratios on this work are
r = a/w = a/(j + t)
And so
r(j + t) = a

As r and j are constants, then, the number of apprentices will rise only as the number of non-resident journeylevel workers temporarily employed here rises. In attempting to limit these workers, the Avalos legislation stands on a concept that is simply wrong.

When faced with unworkable requirements and resulting penalties, contractors will have two choices: to raise their bid prices in anticipation of likely penalties, or to withdraw from the market. Withdrawal becomes more probable as private markets return here or elsewhere in the region. Contractor withdrawal from City-funded work will also result in higher bid prices by reducing competition. Higher bid prices in an era of civic belt-tightening will mean some projects are cut back. Others may simply be eliminated. This, too, will have an effect opposite to what Supervisor Avalos hopes.

Some contractors stand to benefit from the legislation. It is in the short-term interest of any business to reduce its competition. A relative few contractors will benefit from higher bid prices and more opportunities for work. While their workforces may be local, the success of these contractors does not affect the basic mathematics of the legislation.

Additionally, some rough calculations, while not achieving precision, will give a clear picture of the magnitude of difficulty inherent in the legislation.

The City predicts that City-financed construction will produce 59,664 FTE equivalent positions over the next 10 years. An “FTE equivalent” position is 2080 hours of work, or a full 52 weeks of 40 hours.

(The legislation mistakenly calls these “jobs.” Any Trades worker knows that a year of 52 40-hour weeks does not represent a typical Trades “job.” We will examine the distinction in detail below, and then proceed to ignore it in such a way as to favor the Avalos legislation and grant it every possible grace.)

38,408 of these positions will be bunched in the first five of the 10 years; 21,250 will fall in the second five years. The average of annual FTE positions for the second five years is 4,251, but the LMA report does not break these out on an annual basis. If they include swings as wide as the swing of 3,789 FTE positions between 2011 and 2012, they will in one or more of the five years exceed 6,000. These positions include neither Redevelopment Agency work such as Mission Bay, Hunters Point/Candlestick Point, and Schlage Lock, nor work for the Mayor’s Office of Housing, nor work on Treasure Island. All of these will also have local hire requirements independently of the legislation and will impose additional demands on capacity. University of California – San Francisco work will also make local hire commitments. The Transbay Terminal project will likely have local hire commitments, as well. These will add further demand.

Rounding up slightly and not differentiating between the first and second five-year periods, we arrive at an average 6,000 FTE equivalent positions per year in the next 10, just on City-funded work. A 50 percent local hire mandate would mean that 3,000 of these must be San Francisco residents. Again, as labor code specifies a 1:5 ratio of apprentice to journeylevel hours for most trades on public works, approximately 1,000 of the 6,000 FTE equivalent positions will be apprentice positions. The Avalos legislation requires that 50 percent of these, or 500, must be for San Francisco residents. 2,500 of the 3,000 mandated City FTE equivalent positions under the legislation must therefore be journeylevel.

The LMA report gives journeylevel participation of San Francisco residents in City-financed work for 2009-10 as 14 percent and apprentice level participation as 6 percent, for a total of 20 percent. The Brightline/CAA report, drawing on a different data set, gives a total of City resident participation of 24.1 percent, including apprentice participation of 6.4 percent. The Brightline/CAA number approaches the upper end of what has been achieved in San Francisco local hire typically, with rare exceptions in specific and isolated instances but not generally across City work.

For purposes of this calculation, then, let us assume that Trades journeylevel capacity to fulfill a local hire mandate stands presently at about 20 percent; this exceeds what has been demonstrated in either the LMA or Brightline/CAA reports. (This assumption is nonetheless far from incontrovertible, and so I examine it at length below.) Trades capacity to fulfill a local hire mandate would then be 5,000 times 0.2, or 1,000 FTE equivalent journeylevel positions. To fulfill a 50 percent mandate would therefore require 2,500 minus 1,000 FTE equivalent new journeylevel positions, or 1,500.

As noted above, apprenticeships range from three to five years, depending on trade. In recent discussions with Supervisor Avalos’s office, Guillermo Rodriguez of CityBuild stated that on average, apprentices need five years to complete their apprenticeships, regardless of the standard program duration.

Treating City work independently of private and other work, as the legislation affects only City work and as this is the only work for which we have the predictive basis of a 10-year capital plan, the primary means of arriving at the 1,500 new FTE equivalent journeylevel positions will be through the 500 FTE equivalent apprentice positions. They will fill one-third of the 1,500 FTE equivalent journeylevel positions every four or five years, depending on which average we assume to be correct.

Assuming a four-year average for apprenticeship completion, we will then need 12 years to arrive at the 1,500 new FTE equivalent journeylevel positions; assuming a five-year average, we will need 15 years.

(These apprentices will be at all levels, and so the result will be the same if instead of multiplying three cycles of apprenticeship by four or five years we assume that each year a quarter of the apprentices achieve journeylevel status, given a four-year average apprenticeship, or a fifth of them journeylevel status, given a five-year average.)

This calculation takes no account at all of attrition, which is very high in some apprenticeship programs; doing that would raise the total years substantially. Nor does the calculation take into account trade differences. MOST IMPORTANT, THE CALCULATION ASSUMES THAT NONE OF THESE APPRENTICES OR JOURNEYLEVEL WORKERS EVER MOVES OUT OF TOWN. Until and unless the City builds many thousands of units of moderate-income workforce housing, the time needed to achieve 50 percent local hire capacity stretches out indefinitely.

Let us now examine the assumption that 20 percent approximates present journeylevel capacity to fulfill a local hire mandate.

First, it is important to note that an FTE equivalent position does not equate to an actual worker. To make such an equation would require perfect mobility of workers and no differentiation of skills. Frictional unemployment is an ordinary part of life in the Trades. When we are laid off as our work on a project finishes, it is uncommon that we find a call suiting exactly our skills awaiting us in the hiring hall.

Much has been made by local hire advocates and some City officials of the supposed interference of hiring hall procedures in local hire; but even if a laid-off worker could jump any out-of-work list, that worker would still often face wait times between jobs. Employers also ask workers to take time off between phases of a project. Delays and consequent unemployment sometimes occur because of interruptions in material supply. Weather stops some work. Beyond these sources of frictional unemployment, an FTE equivalent position does not take into account that we in the Trades do not have paid vacations but still choose sometimes to take time off. 2,500, 950, 1,550 FTE equivalent positions therefore – taking into account frictional unemployment and elective time off – translate into substantially higher numbers of actual workers.

The LMA report shows 7,855 construction workers resident in the City. Of these, the City employs 1,054 (the 1,377 in the LMA report minus Stationary Engineers and Machinists) directly through Civil Service, for a difference of 6,801. The LMA report does not show how many City resident construction workers are directly employed by other public agencies, including the San Francisco Unified School District, the Community College District, the Golden Gate Bridge District, the Bay Area Rapid Transit District, CalTrans, and the University of California – San Francisco, or by other public agencies outside of the City; nor does it show how many are in long-term employment by hotels, hospitals and apartment complexes. Ignoring these and so giving every possible credence to the Avalos legislation, we can nonetheless subtract from the 6,801 the 1,087 apprentices identified in the LMA report. This gives us 5,714 San Francisco resident construction workers neither employed by the City nor certified apprentices, and so theoretically available for employment on City-financed work.

Of these, if we account for the 20 percent journeylevel capacity we have already acknowledged, and if we equate the 1,000 FTE equivalent positions to actual workers – again to give the Avalos legislation maximum credence, instead of increasing the number to acknowledge frictional unemployment and elective time off – this reduces the available pool to 5,714 minus 1,000, or 4,714.

Who are these 4,714 workers? How many are true journeylevel workers, who in the Trades have hundreds of hours of classroom training and many thousands of hours of work experience? How many are available for City work? Some are day laborers or handypersons, few of whom would qualify as journeylevel. Many others work in the cash economy for any number of reasons and will always avoid the scrutiny of employment on public works. Some are union members whose living situation lets them choose to work a shorter year, or whose health obliges them to work a shorter year. Some are working on state or federal projects or on Redevelopment Agency or Mayor’s Office of Housing work. Many are working in private sector construction, which even in this down economy offers more employment than public work. Many of these in the private sector work for employers who prefer not to bid on public work, or even who will never do so.

The LMA report gives a good indication of who most of these workers are. It shows that 3,968 City resident construction workers who are employed earn less than $20,000 a year. Almost none of these will be true journeylevel workers in either the union or the non-union sector. Few will even be union apprentices. They will instead be candidates for apprenticeship.

In any case, unless the City can positively identify 1,500 true journeylevel workers available and willing to work long term on public works projects from the pool of 4,714, it requires an act of faith to see them there. Public policy affecting many billions of dollars of contracting should not be based on acts of faith.

Much recent discussion of the Avalos legislation has focused on present construction unemployment. Avalos himself has repeatedly described his intent as economic stimulus. The legislation, however, does not sunset when full employment returns, as true stimulus legislation would. It is permanent.

Nonetheless, if we examine current unemployment in the four largest trades in the City, we see that although percentages are high, absolute numbers are not. As of this writing, 243 construction Laborers are unemployed, about 400 journeylevel Electricians, 276 journeylevel Carpenters and about 300 journeylevel Plumbers and Pipe Fitters. These numbers are unfortunately not much different from what they were during the periods from which the LMA and Brightline/CAA reports drew their information.

According to the Brightline/CAA report, about 25 percent of Local 6 electricians are San Francisco residents. It is likely that the percentage is higher among apprentices, lower among journeylevel electricians. Accepting it on its face, though, and seeing no reason that City resident electricians would be unemployed in higher proportions than non-residents, about 100 of the journeylevel electricians now unemployed would be San Franciscans.

We do not have current information on the proportion of City residents among the journeylevel Carpenters of Local 22. In 2000 they made up about 40 percent of the local. That percentage has probably decreased. Using it anyway, we get about 110 City residents among the 276 unemployed Carpenters.

Assuming that the proportion of City residents among the Plumbers and Fitters of Local 38 is the same as we use for the Carpenters – an assumption very generous toward the Avalos legislation – about 120 of the local’s some 300 unemployed journeylevel members would live in town.

Laborers 261 has long been reputed to have a higher proportion of City residents among its members than any other Trades local. Let’s say 95 percent. Of the 243 unemployed construction Laborers, then, about 231 would reside here.

Unemployed journeylevel City residents among other locals, smaller and/or with far larger jurisdictions, will not approach nearly these numbers. The four locals together, under our assumptions (all of which are made in favor of the Avalos legislation) would have 561 City residents seeking work. This is far shy of what is needed to achieve even 40 percent local hire capacity, and even this pool largely dries up when private sector construction rebounds. We are thrown then back on the question of apprenticeship – and on the demonstrated need for non-resident journeylevel workers to advance apprenticeship.

The Avalos legislation is based fundamentally on incorrect concepts. We have been asked to participate in discussions to make it better. We have not been allowed to challenge its concepts. Let’s be clear, then: We can never make it good policy. We can hope only to make it less harmful.