How Policy Took Value Out of Each Prescription
This chapter examines the Portuguese community-pharmacy reforms as an economic mechanism: how changes in regulated wholesale prices, dispensing markups, and the brand–generic mix reduced the value created by each pharmacist’s work. That distinction matters because the later wage response cannot be interpreted until the underlying loss of revenue is clear. A lower wage may reflect a labor-market institution absorbing a real fall in the value pool, rather than an isolated decision by one employer.
The key is to separate three things that are easy to blend together: the activity a pharmacy performs, the margin attached to that activity, and the value available per pharmacist after the workforce is counted. The reforms acted directly on the second component, but their effect appeared in the third. Following that chain shows why a pharmacy could continue filling prescriptions while the economic return to the work fell.
A regulated margin, not an open-ended price
Portuguese community pharmacies operated within a system that constrained both sides of the prescription transaction. They purchased drugs at regulated wholesale prices and dispensed them at prices determined by a regulated markup formula. Pharmacies were therefore not free to raise prices whenever their costs increased or their revenue fell. A change in the rules governing wholesale prices or allowed markups could pass through the sector at the same time, making the shock common to many employers rather than specific to a failing store.
A simple way to see the mechanism is to imagine two versions of the same medicine: a brand-name product and a generic. For each prescription, the pharmacy receives a price related to the regulated wholesale cost and the allowed markup. The pharmacy’s net margin is the part left after the cost of the medicine is accounted for. If the wholesale price falls, the allowed markup is reduced, or a prescription shifts from a higher-margin brand to a lower-margin generic, the margin attached to that transaction can fall even when a pharmacist still performs the work of dispensing it.
The brand–generic mix matters because the margin is not determined only by the number of prescriptions. Suppose a pharmacy fills the same number of prescriptions before and after a reform. If more of those prescriptions are for generic products, and the generic carries a lower regulated margin, the pharmacy can create less value without serving fewer patients. Conversely, a change in the number of prescriptions changes the volume of activity, while a change in the margin changes the value of each unit. Treating those as one variable would make it impossible to tell whether a fall in value came from less work, lower prices, or both.
The study summarizes this relationship through value added per pharmacist. In plain language, that measure combines the value generated by the pharmacy with the number of pharmacists among whom it is counted. The numerator is affected by prescription volume and the net margin per prescription; the denominator changes when the pharmacy employs more or fewer pharmacists. This is why a sector can see value added per pharmacist fall even if total activity does not collapse. A smaller margin reduces the numerator, and a larger pharmacist workforce spreads the remaining value across more people.
Imagine two periods in the same pharmacy. In the first, a relatively high margin is attached to each prescription and a given number of pharmacists share the resulting value. In the second, the number of prescriptions is unchanged, but the margin is lower and the pharmacy adds pharmacists. The work has not vanished, yet the value associated with each pharmacist is lower for two separate reasons. Conversely, a pharmacy could raise value added per pharmacist through greater prescription volume without any improvement in the margin on each prescription. This is why the study tracks both industry value and the workforce denominator: the wage question concerns the value available to pharmacist labor, not just the physical number of transactions.
The measurement is useful precisely because it connects the regulated product market to the labor market, but it is not a complete account of everything a pharmacy does. Community pharmacies also sell non-prescription products and employ technicians, clerical staff, and other service workers. Pharmacists make up roughly one-third of a typical pharmacy workforce. A fall in value added per pharmacist therefore does not mean that every product line, worker category, or service experienced the same change. It is an industry-level indicator of the value associated with the pharmacist workforce, not a direct measure of effort, clinical quality, or the welfare of every employee.
The same caution applies to differences across pharmacies. A pharmacy’s value added can depend on the volume and composition of prescriptions it handles, its staffing, and the product mix available in its location. A regulated formula creates a common exposure to the policy, but it does not force every pharmacy to have the same customers, productivity, or financial position. The central evidence is thus a sector-wide movement, while the later worker outcomes can still vary across firms and people.
The mechanism can be represented without treating every arrow as an equally precise causal estimate:
The lower branch in the diagram is a denominator relationship, not a claim that employment causes value added to fall. When total value is divided across more pharmacist labor, value added per pharmacist is lower unless total value grows at least as quickly. Keeping that accounting point visible prevents a later employment increase from being misread as evidence that the original revenue shock was small.
The reform sequence compressed the value pool
The Portuguese government had already been pursuing measures to lower prescription-drug prices before the deepest crisis adjustments. During the 2010 financial crisis, it introduced administrative changes that reduced maximum pharmaceutical prices and encouraged generic prescribing. Under the bailout-era agreement reached after Portugal requested financial assistance in 2011, the policy package went further. Pharmacists were allowed to substitute generics for brand-name drugs; the reference-price system for generics was changed to draw on relatively low prices in other countries; brand-name prices were automatically reduced when patents expired; and the first generic entering after patent expiry received a specified price treatment. The agreement also changed the margin formulas applying to wholesalers and retail pharmacies.
These changes affected different links in the margin chain. Generic substitution changed the composition of prescriptions. Reference pricing and patent-expiry rules put pressure on the prices of both brand-name and generic products. The wholesale formula changed what pharmacies paid, while the retail formula changed the average markup they could earn. The reforms were therefore not one switch with one effect. They altered the price of inputs, the price or mix of outputs, and the amount retained by the pharmacy between the two.
That also explains why the same reform could have both a public objective and a private labor-market cost. Lower prescription prices and greater use of generics reduce expenditure pressure in the reimbursement system, which is the point of a cost-containment policy. But the pharmacy's margin is part of the revenue from which wages, other operating costs, and returns to owners are paid. Unless the sector can offset the loss through more volume or other products, something in that distribution has to change. The paper does not claim that policy makers selected a wage cut directly; it shows that a smaller value pool created pressure for the firms and bargaining parties that set pay.
The observed margin series shows how quickly the value pool contracted. The authors cite prescription-level calculations in which average net margin per prescription fell from €2.14 in 2010 to €1.78 in 2011, a 17 percent decline. It then fell to €1.37 in 2012, €1.28 in 2013, and €1.24 in 2014. Across 2010–2014, the cumulative nominal reduction was 42 percent. After accounting for a 5.9 percentage-point rise in consumer prices, the paper describes the reduction in the price component of real value added as equivalent to 48 percent.
The chronology gives the result an important qualification. Nearly half of the margin decline occurred between 2010 and 2011, before the later margin-formula changes and institutional provisions targeted by the bailout memorandum were fully in place. The study therefore does not support a simple statement that one clause in the later agreement caused the whole decline. It supports a broader account: several price and product-market changes compressed pharmacy margins over the same general period, while the labor market subsequently adjusted to the smaller value pool.
The fall in value added per pharmacist is consistent with that account, but it is not simply another name for the margin series. The study’s alternative measures of value added per worker moved together during the 2009–2012 reform period and fell by around 35 percentage points from the 2009 level. Between 2012 and 2016, value added per worker rose by about 10 percentage points while value added per pharmacist remained stable. The difference came from staffing: the number of pharmacists at community pharmacies increased by around 20 percent, while the total workforce rose by only about 6 percent, implying a small decline in non-pharmacist employment. After 2016, the two productivity measures again followed more parallel paths.
That divergence is analytically useful. If one looked only at value added per all employees, the 2012–2016 period might appear to show a partial productivity recovery. If one looked only at value added per pharmacist, the recovery would not be visible in the same way. The two measures answer different questions. The first asks how much value is associated with the whole workforce; the second asks how much is associated with pharmacist labor. Because the reforms concerned a sector where pharmacists and non-pharmacists performed different roles, the choice of denominator changes the story one can tell about the labor market.
There is also a data boundary behind the productivity estimates. Value added is directly reported for incorporated pharmacy businesses, but it had to be imputed for sole proprietorships using a model based on sales, employment, and their interaction. That model fits the observed data reasonably well, with an R-squared of 0.88, but an imputed measure is not the same as a directly reported accounting figure. The main conclusion is not dependent on pretending otherwise: the authors compare several measures, and they move in the same direction during the central reform period. Still, the exact magnitude should be treated as an estimate with a measurement component.
The policy-to-margin mechanism also does not tell us how the loss was divided between employers and workers. A pharmacy facing a lower margin might reduce employment, reduce hours, lower wages, accept smaller profits, change its product mix, or combine several responses. The regulated system makes the upstream loss visible, but it does not determine the downstream incidence by itself. That is why the wage-setting rules matter. If sectoral bargaining makes some pay reductions difficult for established workers but permits a cheaper entry schedule, the same common margin shock can be reflected unevenly across cohorts.
Nor does the margin series establish that every pharmacist experienced a 42 percent personal loss. It is an average at the prescription level, and the value-added measures are sector-level or firm-level aggregates. The worker records later show a mixture of incumbent wage erosion, lower entry pay, and changes in the composition of employment. Those outcomes are compatible with the common shock, but they are not mechanically implied by it. The link must be examined through the contract and the worker data rather than assumed from the margin statistic.
The practical lesson from the prescription counter is therefore narrower and stronger than a generic claim that regulation is good or bad. When a government changes a regulated price system, it can reduce the value available to an entire sector without eliminating the underlying service. Employers still need people to fill prescriptions, but the surplus attached to each transaction may be smaller. The next question is who has the bargaining power to preserve their share of that smaller surplus, who can be moved onto a new pay schedule, and how long the adjustment takes.
In Portugal, those questions were answered partly through the sectoral agreement covering pharmacists. The revenue shock created the pressure, but the contract determined the channels through which that pressure could reach different workers. A prescription could remain filled, a pharmacy could remain open, and employment could continue growing while the price of entering the occupation was reset downward. To understand that distribution, we now need to look inside the wage system rather than at the margin alone.