The Causes of Donald Trump’s Authoritarianism

Politics is not like business. Large companies are essentially autocracies, and governments cannot externalise costs, as Trump has done throughout his career. 

It is arguably difficult to find the logic behind Donald Trump’s actions as president. The first few weeks of his presidency have bee marked by turmoil and conflicts with both the media and the courts,  notably when it came to the implementation of his executive order on immigration. The language he and his advisors have been using are reminiscent of the authoritarianism and paranoia displayed by Nixon.

Why is his administration in such turmoil,  while  he was supposed to get things done? There is probably a way to understand Trump’s authoritarian tendencies and his threat to the courts and the media by looking at where he comes from and  the environment in which he came to prominence: a large corporate organisation in an extractive market which essentially works like a dictatorship.  While Donald Trump  sold  his presidential bid on the idea that, as a successful businessman, he would be able to make better deals than traditional politicians, he also seemed to assume that politics and business operate with similar rules. However, there are major differences, especially with the kind of rent-seeking business in which Trump  gained prominence. First, politics operates with many checks and balances, which makes it difficult for a command-and-control type of governing as you see in large companies. Second,  governments can’t externalise costs on a superior actor.  Third, governments cannot choose a line of business and abandon others.

Companies are dictatorships

There are arguably many differences in the way business is governed around the world, but nowhere does it operate like a democracy.  As this article makes clear, large companies are governed like more or less “enlightened dictatorships”. This is especially the case of the United States where mechanisms of employee participation in the strategic decisions of companies is not developed at all. This is very different in Germany, for instance, where the law mandates companies above a certain size to involve employees in decisions affecting not only their employment, but the fate of the company as a whole.  In the United States, in contrast, business leaders expect a clear vertical line of command and compliance at all levels.  Cisco CEO John Chambers, cited in the New York Times, said that

I’m a command-and-control person. I like being able to say turn right, and we truly have 67,000 people turn right.

Things are very different in government. First, the principle of accountability  and division of power are cornerstones of the functioning of government, precisely to prevent the concentration of power in the hands of a single person, as is the rule in companies. Hence, while Donald Trump can fire whoever he wants in his company if they don’t comply with orders, it clearly cannot do so with elected representatives in Congress or judges formally independent from the executive branch. If you want to think of a political form that most resembles the Trump organisation, this would probably be a kingdom where kinship is the primary  mechanism of allocation of power. Trump inherited the company from his father, appointed his children to  head the company, and appointed his son-in-law to a major advisory post in the White House. It is not surprising that he now feels frustrated with an environment that is governed by so different rules and limits on one’s power.

Companies can be bailed out, governments less so

There is an excellent article in the New York Time by Adam Davidson explaining how the economic vision of Donald Trump is entirely shaped by  his experience as a real estate magnate in New York. Manhattan real estate is a good example of what Acemoglu and Robinson  would call an extractive market. Making money in this environment is not really about producing innovative products which  are better or cheaper than the competition, but about securing rents inherited from one’s family or extracted from government via political connections. As Davidson writes

When trying to explain the idea of rent-seeking, I have always used one quintessential example from the United States — a sector in which markets don’t function, in which excess profits are held by a few. That world is Manhattan real estate development. Twenty-three square miles in area, Manhattan contains roughly 854,000 housing units. But there are many more people than that who want to own property there (…) Manhattan real estate development is about as far as it is possible to get, within the United States, from that Econ 101 notion of mutually beneficial transactions. This is not a marketplace characterized by competition and dynamism; instead, Manhattan real estate looks an awful lot more like a Middle Eastern rentier economy. It is a hereditary system. We talk about families, not entrepreneurs. A handful of families have dominated the city’s real estate development for decades: Speyer, Tishman, Durst, Fisher, Malkin, Milstein, Resnick, LeFrak, Rose, Zeckendorf.

 One of the defining features of Donald Trump’s business career, the key of his success, has been his ability to secure rents obtained from politicians for his business ventures,  and his ability to externalise costs on public entities. This started with a 40 year tax break granted by New York City for the renovation of his first hotel,  which has cost $360 million  to taxpayers to date. Trump’s endeavours have so far benefitted from $885 million in tax breaks, grants and other subsidies for luxury apartments, hotels and office buildings in New York. From the little we know about this tax filings, thanks to the intricacies of US tax rules, he was able to file a $900 million loss  in the 1990s, and avoid paying taxes  for up to 18 years. Hence Trump’s business has been to extract political favours for profit, and externalise risk and costs either to the state, or to creditors.

It is very difficult to do something like that in government because there is no superior entity to which costs can be externalised. Arguably, Trump tried to reproduce this logic by saying that Mexico would pay for the wall that he would build at the border, a demand that Mexico flat-out refused. Basically, Trump has evolved in an environment where somebody else can foot the bill for one’s screw ups. In government, beside cases of developing countries  under structural adjustment programs or Eurozone countries in fiscal distress, governments generally have to foot the bill themselves for their own actions.

Governments deal in areas that nobody wants to deal with

Trump’s business career has been marked by constant expansion into new markets to reap new profits, and abandoning old ones. From Queens to Manhattan, to Atlantic City casinos, to Trump University and other ventures. In business, you can close down factories to move them where production is cheaper, and declare bankruptcy on the activity of your casinos. The major difference in government is that it has to do the things that business doesn’t want to do, or cannot do. Barack Obama explained this very well in a speech in October 2016:

Part of government’s job, by the way, is dealing with problems that nobody else wants to deal with. So sometimes I talk to CEOs, they come in and they start telling me about leadership, and here’s how we do things.  And I say, well, if all I was doing was making a widget or producing an app, and I didn’t have to worry about whether poor people could afford the widget, or I didn’t have to worry about whether the app had some unintended consequences — setting aside my Syria and Yemen portfolio — then I think those suggestions are terrific.  That’s not, by the way, to say that there aren’t huge efficiencies and improvements that have to be made. (Government) is not inherently wrecked; it’s just government has to care for, for example, veterans who come home. That’s not on your balance sheet, that’s on our collective balance sheet, because we have a sacred duty to take care of those veterans

Judging from his first weeks, Trump seems to think that the federal government can be operated with the same authoritarian style as the Trump organisation, without regard for checks and balances –  which don’t really exist in the corporate world –  and that the costs of his decisions can be borne by some other actor, most notably other countries. Trump thinks perhaps that he can do with Mexico or China what he did with public entities throughout his career.  He may be up for some hu(uu)ge disappointments.

The network structure of economic expertise in the US and Germany

L’immigration pour garder les femmes à la maison ? Le travail féminin et les politiques d’immigration en Suisse



A paraître dans Panorama, 2017.

Au cours des 60 dernière années, la Suisse a fait massivement appel aux travailleurs et travailleuses étranger-e-s pour satisfaire ses besoins de main d’œuvre. Dans le contexte de reconstruction de l’après-guerre, les travailleurs étrangers, en particulier saisonniers, semblaient être la solution la plus efficace pour augmenter le volume de travail. Pourtant, d’autre pays, Scandinaves en particulier, ont choisi une voie sensiblement différente durant la même période pour augmenter le volume de travail : stimuler l’emploi féminin au travers de l’expansion des crèches, des politiques familiales et des congés parentaux, qui ont été développées beaucoup plus tôt qu’en Suisse. Pourquoi certains pays ont-ils choisi de miser sur l’immigration, alors que d’autres ont choisi de favoriser l’emploi féminin?

Les coûts de l’emploi des femmes

La période qui a suivi la Seconde guerre mondiale jusqu’au milieu des années 1970 a été une période de croissance sans précédent. Pour faire face à une demande croissante de biens et de services, plusieurs pays Européens, comme la Suisse mais aussi l’Allemagne, la France, l’Autriche ou les Pays-Bas ont fait appel de manière massive aux travailleurs étrangers venus des pays du pourtour méditerranéen : Italie, Espagne, Portugal, Yougoslavie, Turquie, Maroc, Algérie. Il faut noter que ces flux étaient en grande partie organisés par les pays d’accueil eux-mêmes, au travers d’accords bilatéraux avec les pays d’origine. Par exemple, l’Union Suisse des Paysans affrétait des trains spéciaux pour amener les saisonniers d’Espagne ou du Portugal vers les exploitations membres en Suisse. Alors que le recours aux travailleurs étrangers dans ces pays était principalement justifié par une pénurie de main d’œuvre dans les pays d’accueil, il faut préciser qu’il s’agissait avant tout d’une pénurie de main d’œuvre masculine. En Suisse, par exemple, le taux d’activité des femmes ne dépassait pas 35% en 1941, et n’a augmenté que très lentement jusque dans les années 1970: il était de 42% en 1971. On peut se demander pourquoi, dans ce contexte, les employeurs et autorités suisses ont préféré faciliter l’immigration de main d’œuvre plutôt que de promouvoir l’emploi des femmes.

Bien entendu, les femmes et les Gastarbeiter de l’après-guerre ne constituent pas des parfaits substituts, en particulier si l’on considère les secteurs dans lesquels ils étaient concentrés : la construction, l’agriculture. Néanmoins, favoriser l’un ou l’autre correspondait aussi à des choix politiques. On peut identifier 4 types de stratégies pour augmenter le volume de main d’œuvre.

Premièrement, on peut favoriser la rationalisation de la production et la mécanisation, qui va permettre de produire plus de biens et de services avec moins de main d’œuvre. Dans certains secteurs, comme l’agriculture, cela pouvait forcer la disparition de beaucoup de petites exploitations, et nécessitait du capital qui était rarement disponible. Deuxièmement, on pouvait augmenter le temps de travail de la main d’œuvre existante. Cette stratégie allait toutefois à l’encontre de la dynamique qui prévalait depuis la guerre, qui allait plutôt vers la réduction du temps de travail.

Troisièmement, on peut tenter d’augmenter le taux de participation des groupes sociaux qui ne prennent pas ou peu part au marché du travail. Le groupe le plus important dans ce contexte était certainement les femmes, et en particulier les femmes en âge d’avoir des enfants. En Suisse, il est intéressant de noter que la participation féminine au marché du travail avait décliné entre 1910 et 1941, de 47% en 1910 à 35% en 1941. Durant cette période, le modèle bourgeois d’un seul salaire (masculin) par ménage s’impose, et l’activité professionnelle des femmes est jugée indésirable tant par les milieux bourgeois que par les syndicats. Dans ce contexte, stimuler l’activité professionnelle des femmes rentre en conflit avec la conception conservatrice de la famille qui prévalait, mais engendre aussi des coûts. En effet, si l’on veut permettre aux mères de travailler, il faut mettre en place un système de crèches, ou une assurance maternité qui, comme on le sait, mettra plus de 60 ans à être réalisée. Toutes ces politiques nécessitent une extension de l’État, des taxes et impôts pour les financer que les milieux bourgeois n’étaient pas prêts à assumer.

La quatrième stratégie, et celle sur laquelle la Suisse a basé l’expansion de son marché du travail, est celle de l’immigration. Cette stratégie comporte l’avantage de ne pas entraîner les dépenses et l’extension de l’Etat social requises par le travail féminin. Les permis saisonniers ne donnaient par exemple pas droit au regroupement familial, ce qui limitait les dépenses liées à l’arrivée de familles. L’immigration n’était par ailleurs pas considérée comme un facteur d’inflation dans la mesure où les travailleurs étrangers étaient censés épargner pour leur retour dans leur pays d’origine plutôt que consommer des biens et services en Suisse. Enfin, “importer” principalement des hommes de l’étranger plutôt qu’encourager l’emploi des femmes suisses était une manière de sauvegarder le modèle familial traditionnel où la femme reste à la maison. Dans les années 1950 et 1960, l’emploi de travailleurs étrangers était perçu comme une politique sans conséquences majeures pour la société suisse dans la mesure où la nature des permis de travail favorisait la rotation des travailleurs. La rotation était censée empêcher les travailleurs étrangers de vouloir et pouvoir s’établir de manière permanente en Suisse. Les droits limités conférés aux travailleurs étrangers les rendaient aussi particulièrement avantageux pour les employeurs, ce qui a créé une demande massive de la part de certains secteurs. Entre les années 1950 et 1990, les nombre de travailleurs étrangers a augmenté de manière importante, et dans des proportions bien plus importantes que dans les autres pays Européens. Le taux d’activité des femmes, par contre n’augmenta que lentement jusque dans les années 1990.

L’exemple alternatif de la Suède

L’expérience de la Suisse est intéressante à comparer à celle de la Suède, qui a pris un chemin bien différent et investi massivement dans l’emploi des femmes dès les années 1960. Si ce pays a également recruté des travailleurs étrangers dans les années 1950 et 1960, en particulier de la Yougoslavie, ces programmes de recrutement ont été de bien moindre ampleur, et ont été interrompus au début des années 1970. Cette différence peut s’expliquer en partie par une importance moindre du modèle familial conservateur, et par des rapports de force politiques plus favorables à l’expansion de l’Etat. En effet, en Suède, le parti Social-Démocrate a exercé une influence décisive sur les politiques publiques de l’après-guerre, gouvernant seul pendant des décennies avec le soutien de puissants syndicats, alors que son équivalent suisse était confiné à une position de minorité au Conseil Fédéral et au parlement. D’une part, la domination du parti social-démocrate constituait un environnement beaucoup plus favorable à l’expansion de l’Etat social nécessaire à l’activation des femmes sur le marché du travail. De fait, les dépenses publiques liées à la maternité en Suède ont été plus de trois fois supérieures à celles de la Suisse jusqu’aux années 2000. Ces politiques ont été promues en masse dès la fin des années 1960. Alors qu’en Suisse l’assurance maternité universelle n’est entrée en vigueur qu’en 2005, la Suède a mis en place un congé parental (qui pouvait être partagé entre pères et mères) en 1974. Par ailleurs, l’activité professionnelle des femmes était activement encouragée, par exemple par le moyen d’émissions de radio. En conséquence, le taux d’activité des femmes a augmenté bien plus rapidement qu’ailleurs. En 1991, alors que le taux de participation au marché du travail des femmes suisses était de 68%, il était de 81% en Suède. Un autre facteur qui a freiné l’emploi de travailleurs étrangers en Suède et que leur emploi était moins avantageux pour les employeurs. En effet, la Suède, n’avait pas l’équivalent des permis saisonniers liés à un employeur ou à un secteur économique, et les syndicats exerçaient un contrôle important sur leurs conditions de travail. De fait les premiers accords de recrutement conclus par la Suède obligeaient les travailleurs étrangers à devenir membres d’un syndicat dès leur arrivée, et la couverture très large des conventions collectives rendaient difficiles pour les employeurs de tirer avantage de salaires plus bas.

Alors que l’immigration en Suisse génère des débats controversés entre acteurs politiques, on oublie souvent que les flux migratoires qui ont affecté la Suisse ont été le résultat de choix politiques. Considérer l’expérience d’autres pays permet de saisir les alternatives possibles et comprendre les chemins qui auraient pu être pris.

Graphique 1: Dépenses sociales liées à l’activité professionnelle des femmes en Suisse et en Suède, 1980-2010




A ranking of European Regions by GDP per capita

Data comes from Eurostat.

Rank Country NUTS Code Name Purchasing power standards (PPS) per inhabitant in percentage of the EU average
1 UK UKI3 Inner London – West 539
2 LU LU00 Luxembourg 266
3 BE BE10 Région de Bruxelles-Capitale / Brussels Hoofdstedelijk Gewest 207
4 DE DE60 Hamburg 206
5 UK UKI4 Inner London – East 204
6 SK SK01 Bratislavský kraj 186
7 DE DE21 Oberbayern 179
8 FR FR10 Île de France 178
9 CZ CZ01 Praha 173
10 SE SE11 Stockholm 172
11 UK UKM5 North Eastern Scotland 164
12 DE DE71 Darmstadt 163
13 NL NL11 Groningen 163
14 DE DE11 Stuttgart 162
15 DE DE50 Bremen 161
16 NL NL32 Noord-Holland 161
17 AT AT13 Wien 158
18 DK DK01 Hovedstaden 157
19 NL NL31 Utrecht 154
20 AT AT32 Salzburg 152
21 IE IE02 Southern and Eastern 150
22 UK UKJ1 Berkshire, Buckinghamshire and Oxfordshire 149
23 FI FI1B Helsinki-Uusimaa 144
24 IT ITH1 Provincia Autonoma di Bolzano/Bozen 144
25 DE DE12 Karlsruhe 140
26 AT AT34 Vorarlberg 139
27 AT AT33 Tirol 138
28 BE BE21 Prov. Antwerpen 138
29 FI FI20 Åland 138
30 DE DE14 Tübingen 136
31 DE DE91 Braunschweig 136
32 DE DE25 Mittelfranken 135
33 DE DEA1 Düsseldorf 134
34 NL NL41 Noord-Brabant 134
35 IT ITC2 Valle d’Aosta/Vallée d’Aoste 133
36 AT AT31 Oberösterreich 132
37 DE DEA2 Köln 132
38 NL NL33 Zuid-Holland 131
39 BE BE31 Prov. Brabant Wallon 130
40 AT AT – ustria 129
41 BE BE24 Prov. Vlaams-Brabant 129
42 RO RO32 Bucuresti – Ilfov 129
43 DE DE23 Oberpfalz 128
44 DE DE26 Unterfranken 127
45 IT ITC4 Lombardia 126
46 DE DE27 Schwaben 125
47 ES ES30 Comunidad de Madrid 125
48 IT ITH2 Provincia Autonoma di Trento 123
49 UK UKD6 Cheshire 123
50 DE DE13 Freiburg 122
51 DE DEA4 Detmold 122
52 DE DE22 Niederbayern 121
53 DE DE92 Hannover 121
54 UK UKI7 Outer London – West and North West 121
55 DE DE30 Berlin 119
56 DE DEB3 Rheinhessen-Pfalz 119
57 DE DEC0 Saarland 119
58 ES ES21 País Vasco 119
59 DE DE73 Kassel 118
60 SE SE23 Västsverige 118
61 IT ITH5 Emilia-Romagna 117
62 AT AT22 Steiermark 116
63 BE BE25 Prov. West-Vlaanderen 115
64 UK UKJ2 Surrey, East and West Sussex 115
65 DE DE24 Oberfranken 114
66 IT ITI4 Lazio 114
67 SE SE33 Övre Norrland 114
68 DE DEA5 Arnsberg 113
69 ES ES22 Comunidad Foral de Navarra 113
70 UK UKH2 Bedfordshire and Hertfordshire 113
71 DK DK03 Syddanmark 112
72 DK DK04 Midtjylland 112
73 UK UKJ3 Hampshire and Isle of Wight 112
74 UK UKK1 Gloucestershire, Wiltshire and Bristol/Bath area 112
75 DE DE94 Weser-Ems 111
76 NL NL22 Gelderland 110
77 DE DE72 Gießen 109
78 DE DEA3 Münster 109
79 NL NL42 Limburg (NL) 109
80 AT AT21 Kärnten 108
81 BE BE23 Prov. Oost-Vlaanderen 108
82 ES ES51 Cataluña 108
83 IT ITH3 Veneto 108
84 PL PL12 Mazowieckie 108
85 DE DEB1 Koblenz 107
86 DK DK05 Nordjylland 107
87 HU HU10 Közép-Magyarország 107
88 NL NL21 Overijssel 107
89 SE SE32 Mellersta Norrland 107
90 FR FR71 Rhône-Alpes 106
91 PT PT17 Área Metropolitana de Lisboa 106
92 SE SE12 Östra Mellansverige 106
93 AT AT12 Niederösterreich 105
94 SE SE22 Sydsverige 105
95 DE DEF0 Schleswig-Holstein 104
96 IT ITC3 Liguria 104
97 IT ITI1 Toscana 104
98 SE SE21 Småland med öarna 104
99 DE DED5 Leipzig 103
100 FR FR82 Provence-Alpes-Côte d’Azur 102
101 IT ITH4 Friuli-Venezia Giulia 101
102 UK UKH1 East Anglia 101
103 ES ES23 La Rioja 100
104 ES ES24 Aragón 100
105 IT ITC1 Piemonte 100
106 NL NL23 Flevoland 100
107 NL NL34 Zeeland 100
108 DE DEB2 Trier 99
109 EL EL30 Attiki 99
110 FI FI19 Länsi-Suomi 99
111 SE SE31 Norra Mellansverige 99
112 UK UKM2 Eastern Scotland 99
113 BE BE22 Prov. Limburg (BE) 98
114 SI SI04 Zahodna Slovenija 98
115 UK UKE2 North Yorkshire 98
116 UK UKI6 Outer London – South 98
117 FI FI1C Etelä-Suomi 97
118 FR FR42 Alsace 97
119 UK UKG1 Herefordshire, Worcestershire and Warwickshire 97
120 ES ES53 Illes Balears 96
121 FR FR62 Midi-Pyrénées 96
122 UK UKF2 Leicestershire, Rutland and Northamptonshire 96
123 DE DED2 Dresden 95
124 FR FR51 Pays de la Loire 95
125 FR FR21 Champagne-Ardenne 94
126 NL NL13 Drenthe 94
127 UK UKD1 Cumbria 94
128 FR FR23 Haute-Normandie 93
129 NL NL12 Friesland (NL) 93
130 UK UKM6 Highlands and Islands 93
131 FR FR61 Aquitaine 92
132 IT ITI3 Marche 92
133 UK UKD3 Greater Manchester 92
134 FI FI1D Pohjois- ja Itä-Suomi 91
135 UK UKE4 West Yorkshire 91
136 UK UKM3 South Western Scotland 91
137 FR FR26 Bourgogne 90
138 UK UKK2 Dorset and Somerset 90
139 UK UKL2 East Wales 90
140 AT AT11 Burgenland (AT) 89
141 FR FR83 Corse 89
142 DE DE40 Brandenburg 88
143 DE DE93 Lüneburg 88
144 DE DEG0 Thüringen 88
145 FR FR24 Centre (FR) 88
146 FR FR52 Bretagne 88
147 IE IE01 Border, Midland and Western 88
148 UK UKH3 Essex 88
149 UK UKJ4 Kent 88
150 BE BE33 Prov. Liège 87
151 DE DED4 Chemnitz 87
152 DE DEE0 Sachsen-Anhalt 87
153 FR FR53 Poitou-Charentes 87
154 IT ITI2 Umbria 87
155 UK UKF1 Derbyshire and Nottinghamshire 87
156 ES ES41 Castilla y León 86
157 MT MT00 Malta 86
158 UK UKG3 West Midlands 86
159 DK DK02 Sjælland 85
160 FR FR25 Basse-Normandie 85
161 FR FR30 Nord – Pas-de-Calais 85
162 FR FR72 Auvergne 85
163 UK UKK4 Devon 85
164 DE DE80 Mecklenburg-Vorpommern 84
165 IT ITF1 Abruzzo 84
166 UK UKC2 Northumberland and Tyne and Wear 84
167 BE BE35 Prov. Namur 83
168 UK UKE1 East Yorkshire and Northern Lincolnshire 83
169 CY CY00 Kypros 82
170 ES ES13 Cantabria 82
171 UK UKD4 Lancashire 82
172 UK UKG2 Shropshire and Staffordshire 82
173 UK UKN0 Northern Ireland (UK) 82
174 FR FR81 Languedoc-Roussillon 81
175 UK UKD7 Merseyside 81
176 UK UKF3 Lincolnshire 81
177 UK UKI5 Outer London – East and North East 81
178 EL EL42 Notio Aigaio 80
179 ES ES11 Galicia 80
180 ES ES12 Principado de Asturias 80
181 ES ES52 Comunidad Valenciana 80
182 FR FR63 Limousin 80
183 CZ CZ06 Jihovýchod 79
184 FR FR41 Lorraine 79
185 ES ES70 Canarias (ES) 78
186 FR FR22 Picardie 78
187 PT PT15 Algarve 78
188 CZ CZ02 Strední Cechy 77
189 FR FR43 Franche-Comté 77
190 FR FRA2 Martinique 77
191 BE BE32 Prov. Hainaut 76
192 BE BE34 Prov. Luxembourg (BE) 76
193 CZ CZ03 Jihozápad 76
194 EE EE00 Eesti 76
195 ES ES63 Ciudad Autónoma de Ceuta (ES) 76
196 PL PL51 Dolnoslaskie 76
197 UK UKE3 South Yorkshire 76
198 BG BG41 Yugozapaden 75
199 IT ITF2 Molise 75
200 LT LT00 Lietuva 75
201 UK UKK3 Cornwall and Isles of Scilly 75
202 ES ES62 Región de Murcia 74
203 UK UKC1 Tees Valley and Durham 74
204 FR FRA1 Guadeloupe 73
205 PL PL41 Wielkopolskie 73
206 PT PT30 Região Autónoma da Madeira (PT) 73
207 ES ES42 Castilla-la Mancha 72
208 IT ITG2 Sardegna 72
209 SK SK02 Západné Slovensko 72
210 HU HU22 Nyugat-Dunántúl 71
211 PT PT20 Região Autónoma dos Açores (PT) 71
212 CZ CZ05 Severovýchod 70
213 CZ CZ07 Strední Morava 70
214 CZ CZ08 Moravskoslezsko 70
215 FR FRA4 La Réunion 70
216 PL PL22 Slaskie 70
217 PT PT18 Alentejo 70
218 IT ITF5 Basilicata 69
219 UK UKL1 West Wales and The Valleys 69
220 ES ES64 Ciudad Autónoma de Melilla (ES) 68
221 SI SI03 Vzhodna Slovenija 68
222 EL EL62 Ionia Nisia 67
223 ES ES61 Andalucía 67
224 PT PT16 Centro (PT) 67
225 EL EL53 Dytiki Makedonia 66
226 PT PT11 Norte 65
227 LV LV00 Latvija 64
228 PL PL63 Pomorskie 64
229 CZ CZ04 Severozápad 63
230 EL EL43 Kriti 63
231 ES ES43 Extremadura 63
232 IT ITF4 Puglia 63
233 PL PL11 Lódzkie 63
234 IT ITG1 Sicilia 62
235 EL EL64 Sterea Ellada 61
236 HU HU21 Közép-Dunántúl 61
237 IT ITF3 Campania 61
238 SK SK03 Stredné Slovensko 61
239 HR HR04 Kontinentalna Hrvatska 60
240 PL PL21 Malopolskie 60
241 IT ITF6 Calabria 59
242 EL EL65 Peloponnisos 58
243 FR FRA3 Guyane 58
244 RO RO42 Vest 58
245 EL EL41 Voreio Aigaio 57
246 HR HR03 Jadranska Hrvatska 57
247 PL PL42 Zachodniopomorskie 57
248 PL PL43 Lubuskie 57
249 EL EL52 Kentriki Makedonia 56
250 EL EL61 Thessalia 55
251 PL PL52 Opolskie 55
252 PL PL61 Kujawsko-Pomorskie 55
253 EL EL63 Dytiki Ellada 54
254 SK SK04 Východné Slovensko 53
255 RO RO12 Centru 52
256 EL EL54 Ipeiros 51
257 EL EL51 Anatoliki Makedonia, Thraki 50
258 RO RO22 Sud-Est 50
259 PL PL33 Swietokrzyskie 49
260 PL PL34 Podlaskie 49
261 PL PL32 Podkarpackie 48
262 PL PL62 Warminsko-Mazurskie 48
263 RO RO11 Nord-Vest 48
264 HU HU33 Dél-Alföld 47
265 PL PL31 Lubelskie 47
266 HU HU23 Dél-Dunántúl 45
267 HU HU32 Észak-Alföld 43
268 RO RO31 Sud – Muntenia 43
269 HU HU31 Észak-Magyarország 42
270 RO RO41 Sud-Vest Oltenia 41
271 BG BG33 Severoiztochen 39
272 BG BG34 Yugoiztochen 39
273 MK MK00 Poranesna jugoslovenska Republika Makedonija 37
274 BG BG32 Severen tsentralen 34
275 RO RO21 Nord-Est 34
276 BG BG42 Yuzhen tsentralen 32
277 FR FRA5 Mayotte 31
278 BG BG31 Severozapaden 30

On the Survival of the Dutch Polder Model: Performance, Populism, Political Economy

To be published (in Dutch) in Beleid en Maatschapij, symposium on “Nog steeds een mirakel? De legitimiteit van het poldermodel in de eenentwintigste eeuw“, ed by Maarten Keune, Amsterdam University Pressnog-steeds-een-mirakel-de-legitimiteit-van-het-poldermodel-in-de-eenentwintigste-eeuw


The totemic status of the Dutch polder model  may have waned somewhat since the publication of Visser and Hemerijck’s “Dutch Miracle” in 1997, but in comparative terms Dutch neo-corporatism – the system whereby trade unions and employers get to decisively shape social and economic policies – still displays a remarkable degree of stability compared with other European countries. Ireland, for instance, another country considered a neo-corporatist miracle in the 1990s and 2000s, has witnessed the virtual collapse of its social partnership model in the aftermath of the financial crisis  (Culpepper and Regan, 2014). While neocorporatism seemed to have undergone a revival in Southern Europe in the 1990s (Hancke & Rhodes 2005), the crisis and the need to cut back spending have led to the marginalisation of employers and trade unions there as well, with the possible exception of Portugal. Austrian corporatism, probably one of the most developed in the world, has seen its legitimacy undermined by populist challengers (the FPö, the Austrian counterpart of the PVV), and scandals involving banks controlled by the trade unions (Afonso, 2013). While Switzerland has also performed well in the aftermath of the crisis, the role of social partners there has also been challenged by party polarisation and the strengthening of populist parties (Haeusermann et al 2004; Afonso & Papadopoulos 2015).

In contrast, while the Dutch economy has also faced serious consequences as a result of the financial crisis, notably a high level of household debt, and its political system  has faced a similar populist upsurge coming from Geert Wilders’s Freedom party, the Dutch Polder model seems to display a good level of resilience in the face of these challenges. Hence, the volume edited by Maarten Keune draws a nuanced picture of the Poldermodel, looking at its economic performance and political legitimacy. Drawing on the contributions of the book, I will address a series of challenges the Dutch border and more all is faced, mainly from our comparative perspective, and seek to answer why it seems to have survived, and the whether it could be exported. I will first draw in assessments of the current state of the Polder model based on the contributions of the book, then address the following challenges: political change and the challenge of populist parties, changes in the economic structure at the level of finance and labour markets, and finally address the potential for the diffusion of the Polder model to other countries in the context of the Eurozone crisis.


As made clear in the chapter by Caelesta Braun, as well as Maarten Keune’s introductory chapter, the legitimacy and resilience of neocorporatism rested on its ability to deliver better economic performance – output legitimacy – which would outweigh its inconvenient in terms of input legitimacy, or its ability to aggregate the preferences of a majority of voters. In the words of Norwegian political scientist Stein Rokkan (1966), in neocorporatism, “votes count, but resources decide”. The Dutch Polder model matched to this idea: social pacts negotiated between trade unions, employers and the government have shaped a number of important reforms in social protection, labour market regulation and wage setting while parliament often merely rubber-stamped agreements. From the point of view of input legitimacy, this is problematic given that only one in six Dutch workers is a member of a trade union.

But is this democratic deficit justified by superior economic performance? As shown by Visser in his contribution, the Dutch economy has been performing well over the last decades in terms of both employment and equality. While employment levels have been among the highest in Europe, especially among women and young people, the Dutch labour market has been astonishingly spared by the rise in inequalities that has spanned all advanced economies since the 1980s. This pattern can be accounted for by the persisting high coverage of collective labour agreements in spite of the dwindling membership of trade unions (outlined in the contribution by Paul de Beer), and the expansion of employment underpinned by the creation of a large number of – mostly part-time – jobs. Hence, the Dutch model is a good example of the “high employment road to low inequality” presented by Kenworthy (2009; Afonso & Visser 2015). Arguably, as Keune makes clear, the Dutch economy has been an outstanding jobs-creation machine, but not a work-creation machine if one considers the huge proportion of part-time jobs in the economy and the total number of working hours performed. Yet, work-sharing via the wide distribution of part-time work – as compared to the concentrated participation on (male) full-time workers that prevails in Southern Europe, for example – is arguably a good solution to achieve low inequality in a political context which doesn’t allow for mass public sector expansion, as in Scandinavian countries. Hence, one possible reason for the survival and resilience of the Polder model is that it has made it possible to deliver both economic performance and provided an effective shield against inequality.

The main output-related argument in favour of the Polder model has been that a particular type of political process involving cooperation between unions and employers has been the cause of the good economic performance of the Dutch economy. The main transmission channel connecting them has been wage restraint: by containing wage increases, the Polder Model has allegedly made it possible to preserve the competitive position of the largely export-oriented Dutch economy, and deliver better employment outcomes. However, this causal relationship can be questioned in a number of respects. One of the most severe criticisms is the one expressed by Ewald Engelen. Rather than an export-driven success, Dutch economic performance has been the result of a hybrid system where the domestic consequences of wage restraint have been compensated by a form of “private keynesianism” propelled mostly by mortgage debt. Hence, high employment happened could have happened despite, rather than because of the Polder Model of wage compression. While wage restraint has dampened domestic demand, mortgage debt has been used to boost it, notably policies such as the Hypotheekrenteaftrek. Hence, the high level of household debt in the Netherlands – Dutch households were the most indebted in the euro area in 2015, with debt representing 283% of household income – may actually be a flipside of the Polder model. This high level of indebtedness is often presented as a major reason for the slow recovery of the Dutch economy after the crisis in comparison with, say, Belgium (FT Alphaville 2016). After the fall of house prices, Dutch households have sought to balance their books and hoarded consumption.


Populism and Electoral Fragmentation

The first challenge faced by the Dutch Polder Model is the increasing fragmentation of the Dutch party system, the weakening of the traditional advocates of corporatist compromises in the legislative arena, and the rise of a populist party claiming to represent the traditional base of trade unions. As emphasised by Keune in his contribution, strong corporatism needs a strong government. This has been especially important when it comes to the regulation of the labour markets via CAOs (explained in the contribution by Verhoeff). In the face of declining trade union membership, political intervention has been a pillar of corporatist compromises by systematically extending collective bargaining outcomes and making them compulsory. Traditionally, the parties that have defended corporatist compromises in the Dutch context have been the PvdA and to some extent the CDA. Now, these parties have faced a dramatic electoral decline over the last decades. In 1993, the CDA and the PvdA represented 66% of parliamentary seats in the Tweede Kamer. After the 2012 elections, this share had fallen to 33%, and projections for the 2017 elections at the time of writing gave 8% and 10% for the PvDA and the CDA respectively. Hence, the channels of transmission of the Polder model into politics have become fairly thin.

The new parties that have emerged or been strengthened in recent years have adopted a generally more critical stance towards neo-corporatist institutions, and had less problems with sidelining trade unions and employers. The “Lenteakkoord” concluded in 2012 between the VVD, CDA, D66, ChristenUnie and GroenLinks was an example of this. One reason is that the extreme proportionality of the Dutch electoral system has facilitated the autonomisation of new middle class constituencies and their move away from the fore of social democracy to new parties (D66, Groenlinks; 50+ and others) with somewhat different socio-economic positions, and the weakening of their organic ties with trade unions. In other countries where the electoral system makes it more difficult for new parties to emerge, these constituencies are kept together, usually under the umbrella of a social-democratic party.

The counterpart of the shrinking of the pro-Polder parties in parliament has been the rise of the PVV as a powerful opposition force with a resolute anti-Polder position. In his Onafhangelijkeidverklaring from 2005, Wilders wrote that the Poldermodel “moet op de helling, zonder pardon, en dat betekent dus het einde van de praat- en overlegpaleizen als de SER en de Stichting van de Arbeid”, as well as “Het afschaffen van het algemeen verbindend verklaren van CAO’s” (PVV 2005). As other populist parties in Europe, the PVV has been very critical of neo-corporatist institutions perceived to work like a political cartel from which it is mostly excluded. Dismantling it is a way to undermine the cosy relationship between mainstream parties and interest groups, and especially trade unions. Interestingly, trade unions typically represented the socio-economic clienteles that the PVV is now mostly claiming, namely older, low-and-middle skilled working class workers. Weakening their influence can also be part of a strategy to become the only voice of this constituency. In 2011, for instance, the PVV proposed to give voting rights to non-union members in the negotiations of collective labour agreements as a way to dilute the influence of labour organisations (BNR 2011)

This being said, the increasing fragmentation of the party system may not necessarily mean the end of the Polder model. Indeed, as examples in other countries show, the polarisation of politics may actually constitute an opportunity for social partners because of their continuing ability to reach compromises when political parties can’t (Afonso 2013). Hence, in the perspective of a multi-party government in which agreements may be difficult to reach, readymade solutions by social partners may still be able to make their way into policy.


Political-Economic Change

The last important challenge for the Poldermodel is the change in the economic structure. This mainly relates to the evolution of the labour market and its attainment of a potential ceiling and the increasing financialisation of the Dutch economy, especially when it comes to the pension system (explained in the contribution by Van der Zwan).

The first challenge has to do with the ability of the Polder model to deliver superior economic performance, especially when it comes to women. In many respects the “jobs miracle” that took place in the Netherlands in the 1980s and 1990s was based on the creation of part-time jobs mainly taken up by women. This expansion was fairly spectacular, but it was mainly a catching-up phenomenon in comparative terms because female employment in the Netherlands started at very low levels. In fact, The Netherlands had the lowest level of female labour force participation in Europe 1971, nearly 20% below the OECD average. The publication of the Dutch Miracle in 1997 corresponded to this level just about reaching the OECD average. Since then, the potential for increasing female labour force participation has shrunk, and the rate of increase between 1996-2015 and 1985-1995 has halved. Now it is unclear which potential source of labour can be used to increase the labour force. If the output legitimacy of the Polder Model was based on mass job creation, it may now be running out of steam.

The second aspect relates to the increasing financialisation of the Dutch economy, the in-built volatility this introduces, and how it changes the preferences of the actors involved. First, the compromises that underpinned the Polder Model were premised on the ability of domestic actors to exert some control over economic parameters via economic policy choices. Second, it was based on a class compromise between actors with clear interests, namely employers wanting wage restraint and trade unions wanting jobs, better purchasing power and/or social protection. As outlined in van der Zwan and Engelen’s contributions, these assumptions have been severely challenge by the tremendous increase in the level of financialisation of the Dutch economy. In 1995, stocks traded represented 25% of Dutch GDP. By 2000, they represented 150%, and 183% in 2007, with very high levels of volatility, notably after the bubble and the global financial crisis. This factor has challenged some basic tenets of the Polder model in two major ways. First, the heavy level of involvement of pension funds jointly managed by social partners in financial markets has blurred the interests of capital and labour. The future income of workers via pension benefits is tied to investment decisions which may go against the interests of other workers, for instance by favouring short-term profits over long-term investments. In a nutshell, workers have become capitalists by procuration. Secondly, and most importantly the high level of volatility and uncertainty brought in by financialisation has made corporatist compromises extremely vulnerable to events happening in financial markets, which are globalised. Hence, it has become more difficult for corporatist actors to make credible commitments – eg. about pension benefits – since revenues are largely beyond their direct control, and may undergo high levels of volatility.

In sum, while the Polder model has indeed displayed a fairly high degree of resilience, it hasn’t been immune to broader political and economic developments which could possibly fundamentally transform it in the long term. While it is unlikely to disappear, it may however be deprived of much of its regulatory capacity in a context of financialisation, and the economic “Miracle” which attracted wide international attention is going to be difficult to sustain, let alone renew. Keune’s edited volume convincingly recognises its successes, but also convincingly highlights its many hidden vectors of instability.


Afonso, Alexandre, and Yannis Papadopoulos. 2015. “How the Populist Radical Right Transformed Swiss Welfare Politics: From Compromises to Polarization.” Swiss Political Science Review 21(4): 617–35.

Afonso, Alexandre. 2013. Social Concertation in Times of Austerity. Amsterdam: Amsterdam University Press.

BNR. 2011. Jacht op vakbonden is geopend.

Culpepper, P. D., and A. Regan. 2014. “Why Don’t Governments Need Trade Unions Anymore? The Death of Social Pacts in Ireland and Italy.” Socio-Economic Review 12(4): 723–45.

Afonso, A and Visser, J. 2014. “The Liberal Corporatist Road to High Employment and Low Inequality? The Dutch and Swiss Social Models in the Crisis, in Dølvik, Jon Erik, and Andrew Martin. 2015. European Social Models from Crisis to Crisis: Employment and Inequality in the Era of Monetary Integration. Oxford: Oxford University Press.

Financial Times. 2016. Why is the Netherlands doing so badly? (January 4, 2017).

Hancke, B. and Rhodes, M. 2005. “EMU and Labor Market Institutions in Europe: The Rise and Fall of National Social Pacts.” Work and Occupations 32(2): 196–228.

Häusermann, Silja, André Mach, and Yannis Papadopoulos. 2004. “From Corporatism to Partisan Politics: Social Policy Making under Strain in Switzerland.” Swiss Political Science Review 10(2): 33–59.

Kenworthy, Lane. 2009. “The High-Employment Route to Low Inequality.” Challenge 52(5): 77–99.

Rokkan, S. 1966. “Norway: Numerical Democracy and Corporate Pluralism.” In Political Oppositions in Western Democracies, ed. R. A. Dahl. Yale: Yale University Press.

Partij voor de Vrijheid. 2005 “Onafhangelijkheidsvreklaring (January 4, 2017).

Book review: How a strong Communist party led to less social protection in Portugal, and its absence to more protection in Spain

9780190245474The Left Divided: The Development and Transformation of Advanced Welfare States By Sara E. Watson, Oxford: Oxford University Press, 2015. ISBN: 9780190245474, 376 p (review to appear in Political Studies Review, autumn 2017).

One of the most widespread assumptions in welfare state and comparative political economy research is that the strength of the left is the major driver of redistributive policies in advanced industrialized countries. However, most of the classic works in the field have assumed that the left constitutes an homogeneous bloc. Scandinavian countries, where social-democratic parties have been closely connected to powerful trade unions, have constituted the main empirical basis of important works advocating this position, such as those of Esping-Andersen (1992) or Korpi (1983).

In her impressive comparative analysis of Spain and Portugal, Sara Watson shows that this assumption of homogeneity within the left is more the exception than the norm. In many countries, the left is divided across the party and industrial relations system between a moderate and a radical left, and these divisions and competition patterns among left-wing parties can yield counterintuitive effects on inequality and redistribution. Hence, she shows how Portugal, where a powerful Communist party was able to control the trade union movement during the transition to democracy, developed a more unequal and liberal welfare model in the 1970s and 1980s. Spain, meanwhile, where the transition was controlled by the right, established a system with a higher level of social protection for workers in the aftermath of transition. This unexpected outcome, Watson argues, was the result of partisan strategies. Faced with a strong far-left competitor, the Portuguese moderate Socialist Party adopted a deliberate strategy to weaken the Communists and their trade union allies organizationally, resulting in a higher level of exposure of Portuguese workers to market forces. In the absence of comparable patterns of intra-left competition, no such incentives for liberalization existed in Spain. Accordingly, this political process shaped labour market outcomes in both countries up to now, with higher social protection but higher unemployment in Spain, and low unemployment but more poverty and inequality in Portugal.

To support this clever theoretical argument, Watson draws on an impressive comparative historical analysis connecting the intricacies of intra-left competition with changes in industrial relations, employment protection and income compensation. Combining theoretical innovation with an impressive comparative-historical analysis, this is by far the best account of the development of welfare capitalism in Spain and Portugal after democratisation, and its insights will be relevant for scholars working on a much broader range of countries.


Esping-andersen, G. 1992. Politics Against Markets. Princeton: Princeton University Press.

Korpi, W. 1983. The Democratic Class Struggle. London: Routledge.


Democracy in Cuba, 1900-2010


The V-Dem dataset is a huge dataset that compiles many indicators to measure political institutions, dating back to 1900. With the death of Fidel Castro, I tried to see what the indicators said about the evolution of democracy in Cuba. It shows a tremendous decline after 1950. Interestingly this started before Fidel Castro, during the dictatorship of Fulgencio Batista. When Castro took power, the democratic index stayed low, essentially continuing the type of autocratic rule pursued by Batista with a different ideological bent.