In the past several years, government policies have resulted in a dramatic change in the composition of Cuba’s labor force.
- Employment in the state sector plunged by 587 thousand, In the period 2011-2014, about 16% of the state workforce. (See rightmost column of Table 1 and the figure at the end of the paper).
- The vast majority of the employees released by the state were absorbed by the non-state sector, where employment surged by 572 thousand.
- Open unemployment remained basically unchanged during the period 2011-14 as whole, and the decline in total employment and labor force was small. There was a large increase in the estimated number of discouraged workers, implying that some of the employees released from the state sector decided to quit the labor force, at least temporarily. But this was largely offset by an upsurge in the population of working age. The puzzling coincidence of these two developments is examined further below.
- Most of the increase in non-state employment occurred in the private sector; employment in the (overwhelmingly agricultural) cooperatives rose only 15 thousand). 
- Within the private sector, employment rose sharply in both the self-employment and the “other private” categories; the latter consists primarily of small private farmers.
The unprecedented change in the composition of employment is shown in Table 2. The share of state employment dropped by 11 ½ percentage points from 2010 to 2014 and the non-state share went up as roughly as much. Virtually all of the increase in the non-state sector corresponded to the private sector, where the employment share reached a peak of almost 28 % in 2014—a level unseen in Cuba for more than four decades, although it remains small by international standards. Self-employment and “other private” jobs increased substantially, both reaching new peaks.
Two major phases of the re-allocation process should be distinguished. The major part of the labor transfer occurred in 2011-12, the first two years of the program, when 483 thousand employees left the state sector and 401 thousand took up non-state jobs. Open unemployment went up by almost 50 thousand in that period and both employment and the labor force declined.
Developments in the period 2013-14 were more complicated. The transfer from the state to the non-state sector slowed considerably, but it did not stop: just over 100 thousand employees left the state sector, and non-state employment rose by 172 thousand. The number of unemployed fell by 40 thousand and the labor force actually rose. However, the estimated number of discouraged workers surged, largely offsetting an unusually large (252 thousand) increase in the population of working age (PWA), which far exceeded the rise in the labor force. A possible interpretation of this seemingly strange result could that the PWA rose for strictly exogenous (demographic) reasons, while the number of discouraged workers jumped because some of the employees released from the state workforce decided to withdraw, at least temporarily, from the labor force, as they found it difficult to find immediately an appropriate job—one in which their age and lack of experience would not be an overly severe handicap.
The remaining question is why the PWA increased so sharply in 2013-14? In an accounting sense, the answer would seem to be a combination of two factors: a rise in the resident population (following a 7-year decline); and a huge drop in the population not of working age. Perhaps the demographers (or the statisticians at ONEI) could explain that one.
The rise in the private sector share of employment was widespread geographically. Among the 8 provinces for which sufficient data was available, all except Guantanamo registered a rise in the private share in 2011-14, with the largest increases in Matanzas (21 percentage points), Havana (20), Holguin (17) and Cienfuegos (13); Naturally, Havana registered the largest absolute increase—just over 200 thousand or 45% of the country-wide rise in employment. In almost all the provinces, both self-employment and other private jobs increased, with the latter dominating in Holguin, Cienfuegos, Granma, Matanzas and Camaguey, and the former in Havana and Santiago. Among the few provinces that publish the composition of the “other private” category, most reported that private farmers had the largest share in this sector, with mixed enterprises, associations, foundations, and foreign enterprises coming well behind.
The breakdown of employment by economic sector provides some confirmation of the results presented above. Of course most of the categories under that breakdown include a mix of both state and non-state sub-components, and therefore an exact comparison with the institutional breakdown examined above is not possible. It is interesting to note, however, that the sum of employees in four categories that are known to include a large proportion of non-state producers (commerce, restaurants & hotels, construction, transport & storage, and agriculture & fisheries) increased by 172 thousand from 2010 to 2014 while jobs fell by 186 thousand in all other sectors. This implies that the non-state share in all other categories combined must have increased substantially. More than half of the rise in non-state workers was in commerce, hotels & restaurants; and more than three fourths of the fall in state jobs was in the category of ‘communal & social services’.
In other labor market developments there was a 24% increase in state sector wages in 2104, compared with 1% in 2013 and 2 ½ % in 2012. Awards were particularly large in public health (61%), perhaps reflecting salary adjustments and new contracts with Brazil and some Middle Eastern countries); commerce and repairs (53%); and mining (44%). They were very small (1.3%) in education. Perhaps not coincidentally, the unusually large wage increase in 2014 coincided with an exceptionally strong (21%) growth of the money supply (M2A), which compares with an annual average of 6% in the four previous years.
In conclusion, the inter-sectorial transfer operation appears to have been a success, even though the authorities’ original target of releasing 1.3 million employees—a target which may have been overly ambitious—was not achieved. The size of hidden unemployment in the state sector has been considerably reduced, private sector employment was boosted, job search encouraged, and the result will be an increase in labor productivity over the medium term. The transfer was achieved at very little cost in terms of open unemployment but the estimated rise in discouraged workers was very large, particularly in the late phase of the program. Given the strong sensitivity of these workers to changes in real wages, however, it is possible that this rise will be reversed in the future, if wage increases continue to be strong. Perhaps a more difficult question is whether there is enough room in the private sector for further sizeable transfers of manpower given present restrictions. If the answer is negative the authorities may have to significantly expand the list of profession open to private employment, which may require a re-consideration of the state monopoly on activities like health, education, culture and sports.
 There is no published information on the number of discouraged workers in Cuba. That number, which was estimated in this paper by the difference between the potential and the actual labor force, was found to be highly cyclical and strongly correlated with the real wage rate.
The potential labor force is the hypothetical level of the workforce expected to be available if labor resources were fully utilized. It is calculated by evaluating the actual labor force at the maximum historical level of the participation rate.
 For all the talk about non-agricultural cooperatives, the number of workers reported ONEI under this heading is minuscule.
 This is illustrated in the figure, where the positive portions of the histograms represent the factors that absorb the decline in state employment. In each histogram, the difference between the negative and the components is equal to the fall in state employment.