Effectiveness and efficiency of state aid for new broadband networks: evidence from OECD member states

Figures & data

Table 1. Major national broadband/NBN programs introduced in 34 OECD states until 2018.

State	National broadband/NBN program names^a	National NBN targets^b	National NBN funds^c	(%/Focus) Rural areas^d
European OECD states
AUT	ICT Master Plan – 2007; Breitband 2013	2020: 99% of households with 100 Mbit/s	Yes	100%
BEL	Digital Belgium	2020: 100% of households with 30 Mbit/s	No	–
CHE	Digital strategy	2020: 100% of municipalities with 30 Mbit/s	No	–
CZE	Digital Czech Republic; National Plan for Develop. of NGN	2020: 100% of population with 30 Mbit/s population and a coverage of at least 100 Mbit/s for 50% of households	Yes	Yes
GER	IKT 2020; Digitale Agenda	2018: 100% of households with 50 Mbit/s	Yes	94.3%
DNK	Better broadband coverage	2020: 100% of households and businesses with 100 Mbit/s	Yes	100%
ESP	Broadband strategy; Digital Agenda	2020: 100% of population with 30 Mbit/s population and a coverage of at least 100 Mbit/s for 50% of households	Yes	60%
EST	Information Society Development Plan; EstWin Project	2020: 100% of households with 30 Mbit/s	Yes	100%
FIN	Broadband implement. Plan; ubiquitous information society	2015: 99% of households, businesses and public offices with 100 Mbit/s	Yes	100%
FRA	Digital France 2012; France très haut débit	2022: 100% of households, businesses and public offices with 30 Mbit/s	Yes	100%
GRC	National Broadband Plan Next Generation	2020: 100% of households with 30 Mbit/s	Yes	75%
HUN	Digital Literacy Action Plan; National Infocommunication Strategy	2018: 100% of households with 30 Mbit/s 2020: 50% household penetration with 100 Mbit/s	Yes	71.4%
IRL	Next Generation Broadband; national broadband strategy	2020: 100% of households with 30 Mbit/s	Yes	100%
ISL	e-nation	2020: 99.9% of households and businesses with 100 Mbit/s	No	–
ITA	Broadband Internet for all; Strategy for next gen. access network	2020: 100% of households with 30 Mbit/s; 100% of businesses and 85% of population with 100 Mbit/s	Yes	75%
LUX	National Strategy for superfast broadband	2020: 100% of households, businesses and public offices with 1 Gbps/500 Mbit/s	Yes	No
NLD	Broadband policies; Digital Agenda	2020: 100% of households with 30 Mbit/s	Yes	100%
NOR	Digital Agenda; Information Society	2020: 90% of households with 100 Mbit/s	Yes	Yes
POL	National broadband plan; Development of the Information Society	2020: 100% of households and businesses with 30 Mbit/s and 50% of households with at least 100 Mbit/s	Yes	75%
PRT	Agenda Digital; connecting Portugal	2020: 100% of population with 30 Mbit/s population and a coverage of at least 100 Mbit/s for 50% of households	Yes	100%
SVK	Digital growth & next gen. infrastructure; Information Society Strategy	2020: 100% of households with 30 Mbit/s	Yes	Yes
SVN	Next gen. broadband networks; Broadband Network Development Strategy	2020: 96% of households with 100 Mbit/s and the remaining 4% with 30 Mbit/s	Yes	Yes
SWE	Completely Connected Sweden; Rural areas program	2025: 98% of households and businesses with 1 Gbit/s	Yes	100%
GBR	NGN Deployment Plan; Superfast Broadband Future	2020: 95% of households and businesses with 25 Mbit/s	Yes	100%
Non-European countries
AUS	National Broadband Network	2020: 90% of households and businesses with 50 Mbit/s	Yes	Yes
CAN	Connect to Innovate Connecting Rural Canadians	2021: 90% of households and businesses with 50 Mbit/s	Yes	Yes
CHL	Digital action plan Agenda Digital	2022: 90% of households with 10 Mbit/s	No	–
ISR	Communicat­ion Initiative	2022: 100% of population with 30 Mbit/s	No	–
KOR	IT Korea 5 Future Strategies; Plan for broadband convergence; Giga Internet Promotion	2017: 90% of urban areas with 1 Gbit/s and 99% of households with 100 Mbit/s	Yes	Yes
JPN	U-Japan policy; new strategy in ICT; Creation of most advanced IT nation	2022: Reduce the number of households not covered by FTTH to 180.000	Yes	Yes
MEX	Connectado progamme	2024: 92.2% of the population with 4 Mbit/s downlink per 1 Mbit/s during the network peak hour	No	-
NZL	Ultra-fast broadband program; rural Broadband Initiative	2025: 99% of households with 50 Mbit/s and the remaining 1% with 10 Mbit/s	Yes	Yes
TUR	Information Society Strategy	2020: 50% of households with 100 Mbit/s 2023: 100% of them with 100 Mbit/s	No	–
USA	National Broadband Plan; American Recovery and Reinvestment Act	2020: 80% of households with 100 Mbit/s	Yes	Yes

Sources: ^aOECD (2018, 41); OECD (2011, 24–28), ^bOECD (2021, 26–27); ^ccountry-specific funding sources are cited in Table A2 in the Appendix;^ddata on shares of rural areas funded in European states is taken from WIK et al. (2020, 21); if these data were not available, we indicate whether the public funding focus was based on rural (‘white’) areas or not (‘Yes’/‘No’) based on the sources cited in Table A2.

Table 2. Summary statistics.

Variable	Obs.	Mean	Std. Dev.	Min.	Max.
Dependent var.:
FiberCov	512	0.78	0.74	0	2.86
Main var.:
StateAid	512	0.53	0.50	0	1
Control vars.:
FixMobSub	512	0.60	0.19	0.08	0.90
ShareDSL	512	0.39	0.22	0.01	1.59
CommExp	512	2.84	0.69	1.39	4.94
GDPpc	512	40.00	15.67	10.87	120.98
Netflix	512	0.37	0.48	0	1
IntUsers	512	74.68	16.97	14.58	99.01
TeleServP	480	98.79	16.80	48.80	258.70
AgeDepRatio	512	50.38	5.67	36.21	68.28
RuralPop	512	0.23	0.12	0.02	0.49
IntRate	512	3.71	2.87	−0.49	22.50
Wages	512	41.08	19.99	7.70	97.20
ShareApprt	512	41.48	15.68	6.91	70.16
AccessReg	327	0.42	0.49	0	1
AccessPrice	204	11.30	3.92	5.34	31.30
Instrumental vars.:
LftWng	512	39.25	12.11	9.80	69.09
GovExp	512	42.69	8.16	19.12	65.03
FiberCov31	512	0.75	0.59	0.01	1.67
StateAid31	512	0.50	0.21	0.12	0.73

Notes: We used linear interpolation to close the gaps in our time series. Some 0.88% of the observations were calculated this way. Missing values are not related in any systematic pattern with regard to NBN deployment or state aid programs.

Figure 1. NBN investment and adoption household shares (OECD mean values for 2003–2019). Source: Own calculations based on data from FTTH Council Europe.

Table 3. Main estimation results.

Dep. Var.: FiberCov(t)	(1)	(2)	(3)	(4)	(5)
FiberCov(t−1)	0.440*** (0.112)	0.562*** (0.096)	0.508*** (0.137)	0.628*** (0.097)	0.676*** (0.084)
StateAid(t)	0.220*** (0.084)	0.279*** (0.077)	0.345*** (0.108)	0.249*** (0.060)	0.356*** (0.087)
FixMobSub(t)	0.903*** (0.310)	0.574** (0.238)	0.611 (0.397)	0.717*** (0.224)	0.496*** (0.187)
AgeDepRatio(t)	0.019* (0.011)	0.012* (0.007)	0.015* (0.008)	0.007 (0.008)	0.002 (0.007)
GDPpc(t)	0.010* (0.006)	0.007 (0.005)	0.008 (0.006)	0.005 (0.004)	0.004 (0.004)
RuralPop(t)	−5.141*** (1.238)	−4.376*** (1.200)	−4.682*** (1.416)	−3.325*** (1.268)	−3.110** (1.326)
IntRate(t)	−0.015** (0.007)	−0.015** (0.006)	−0.016** (0.006)	−0.020** (0.009)	−0.021** (0.009)
F-test (2SLS)/χ^2 (GMM) (p-value)	0.000	0.000	0.000	0.000	0.000
AR(2) test (p-value)	0.802	0.727	0.668	0.729	0.625
Hansen test (p-value)	0.110	0.426	0.340	0.767	0.115
# of instruments	18	15	11	42	26
# of groups/clusters	32	32	32	32	32
# of observations	480	512	512	480	512

Notes: Columns (1), (2), and (3) are based on a 2SLS estimator (Anderson and Hsiao 1981), which uses various IV-style internal and external instruments that differ only in the lag structure. Columns (4) and (5) are based on a one-step GMM-diff. estimator (Arellano and Bond 1991), which additionally uses GMM-style instruments for the lagged dependent variable. Standard errors in parentheses are clustered at the group (i.e. country) level and robust to arbitrary forms of heteroscedasticity.

*p < 0.10.

**p < 0.05.

***p < 0.01.

Table 4. Tests and robustness checks.

Dep. Var.: FiberCov(t)	(1)	(2)	(3)	(4)	(5)
FiberCov(t−1)	0.687*** (0.136)	0.690*** (0.110)	0.605** (0.257)	0.582*** (0.141)	0.287 (0.254)
StateAid(t)	0.279** (0.137)	0.239** (0.113)	0.374*** (0.114)	0.209* (0.113)	0.347** (0.152)
FixMobSub(t)	0.354** (0.151)	0.341** (0.148)	0.537** (0.261)	0.643** (0.269)	0.989** (0.482)
AgeDepRatio(t)	0.002 (0.007)	0.008 (0.006)	0.003 (0.007)	−0.011 (0.013)	0.008 (0.017)
GDPpc(t)	−0.000 (0.004)	0.001 (0.004)	0.005 (0.005)	0.016** (0.007)	0.013 (0.008)
RuralPop(t)	−2.961* (1.709)	−2.561* (1.476)	−3.337** (1.559)	−4.566*** (1.351)	−7.081*** (2.187)
IntRate(t)	−0.020** (0.009)	−0.020** (0.009)	−0.021** (0.009)	−0.021* (0.012)	−0.019** (0.009)
TimeTrend(t)	0.009 (0.021)	0.018 (0.016)
TimeTrend^2(t)		−0.001 (0.000)
FiberCov(t−1) * StateAid(t)			0.059 (0.228)
AccessReg(t)				0.019 (0.077)
AccessPrice(t)					−0.004 (0.004)
χ^2-test (p-value)	0.000	0.000	0.000	0.000	0.000
AR(2) test (p-value)	0.692	0.818	0.510	0.185	0.142
Hansen test (p-value)	0.109	0.777	0.129	0.374	0.284
# of instruments	27	28	27	27	19
# of groups/clusters	32	32	32	27	21
# of observations	512	512	512	327	204

Notes: ALL regressions use a one-step GMM-diff estimator (Arellano and Bond 1991) and extend the results from Table 1, column (5) by including additional explanatory variables, as listed in the table. Standard errors in parentheses are clustered at the group (i.e. country) level and robust to arbitrary forms of heteroscedasticity.

*p < 0.10.

**p < 0.05.

***p < 0.01.

Table 5. State aid expenditures on NBN.

Country	State aid per capita (US$)	Period	# of years	State aid per capita per year (US$)
Australia	1721	2003–2019	17	101.24
Austria	125.5	2003–2018	16	7.84
Belgium	n/a	n/a	n/a	n/a
Canada	41.4	2002–2019 (except 2013–2015)	15	2.76
Czech Rep	1.2	2003–2018	16	0.08
Denmark	8.4	2014–2019	6	1.41
Finland	30.3	2003–2018	16	1.89
France	214.8	2003–2018	16	13.43
Germany	87.3	2003–2018	16	5.46
Greece	44.2	2003–2018	16	2.76
Hungary	n/a	n/a	n/a	n/a
Iceland	n/a	n/a	n/a	n/a
Ireland	40.2	2003–2018	16	2.51
Italy	145.4	2003–2018	16	9.09
Japan	11.6	2006–2010	5	2.33
Korea	362.0	2002–2012	11	32.91
Luxemburg	n/a	n/a	n/a	n/a
Netherlands	1.5	2003–2018	16	0.09
New Zealand	270.0	2009–2019	11	24.55
Norway	60.5	2002–2014	13	4.65
Poland	17.7	2003–2018	16	1.11
Portugal	10.1	2003–2018	16	0.63
Slovak Rep	21.0	2003–2018	16	1.31
Spain	34.3	2003–2018	16	2.14
Sweden	7.0	2003–2018	16	0.44
Switzerland	4.3	2007–2019	13	0.33
Turkey	n/a	n/a	n/a	n/a
UK	59.5	2003–2018	16	3.72
USA	129.9	2009–2019	11	11.81
Estonia	16.8	2003–2018	16	1.05
Israel	n/a	n/a	n/a	n/a
Slovenia	75.5	2009–2019	11	6.86
Average	136.21		14.2	9.32

Sources: own calculations based on Feasey, Bourreau, and Nicolle (2018) and sources cited in Table A2.

Table A1. Variable descriptions and sources.

Variable	Description	Source^a
	Dependent variable
FiberCov	Household-weighted number of ‘homes passed’ by relevant FTTx technologies: Fiber to the Cabinet (FTTC) based on VDSL technology, Fiber to the Node (FTTN) based on DOCSIS 3.0 technology, Fiber-to-the Building (FTTB), and Fiber to the Home (FTTH) technologies. In FTTH/FTTB scenarios, the final connection to the subscriber is optical fiber and terminates inside the premises, on an external wall of the subscriber’s premises, or no more than 2 m from an external wall. ‘Homes passed’ is the total number of premises. The premises is a home or place of business.	FTTH Council Europe^b, own research
Main variable of interest: State aid
StateAid	Dummy variable that takes the value 1 if there is a national state aid program implemented, supporting NBN deployment in a given year in the respective country (otherwise zero).	Own research (see Table A2)
Competition and regulation controls
ShareDSL	Absolute number of wired broadband subscribers using DSL technology divided by absolute number of standard analog telecommunication access lines. DSL lines include bandwidth levels > =  256 kbit/s but exclude leased lines.	OECD Broadband Database^c
FixMobSub	Share of the total number of mobile-cellular broadband subscriptions to the total number of mobile-cellular broadband subscriptions and total number of active fixed broadband subscriptions.	MarketLine Advantage^d
AccessReg	NBN wholesale access regulation including all remedies imposed on FTTx infrastructures of market-dominant operators. NBN access regulation is measured as a binary indicator, which is equal to 1 in years in which at least one of the available access remedies are in force in a given OECD member state (otherwise zero).	Own research (Briglauer, Cambini, and Grajek 2018)
AccessPrice	Average access price for full local loop unbundling in € which is calculated as the regulated monthly fee plus the regulated fixed connection fee distributed over three years.	Own research (Briglauer, Cambini, and Grajek 2018)
Other revenue controls
CommExp	Percentage of total household expenditure used for ICT goods in the respective year.	MarketLine Advantage^d
GDPpc	Gross domestic product per capita converted to thousands of US dollars using current prices and current purchasing power parities.	OECD National Accounts Database^e
AgeDepRatio	Ratio of dependents (people younger than 15 or older than 65) per 100 working-age individuals.	World Bank^f
IntUsers	Internet users are people with access to the World Wide Web network, measured in percentage of the population.	ITU^g
TeleServP	Index (2010 = 100) putting in relation the prices of telecommunications services in different years.	Euromonitor^h
Netflix	Dummy variable which takes on value 1 if Netflix streaming services were available (otherwise zero).	Own research
Cost controls
RuralPop	Rural population measured as the percentage of the total population living in rural areas, calculated as difference between total population and urban population (as defined by national statistical offices).	World Bank World Development Indicators^i
IntRate	Long-term interest rates refer to government bonds maturing in ten years. The interest rates are generally measured as averages of daily rates, expressed as a percentage.	OECD Finance Database^j
ShareApprt	Total number of households by type of dwelling (apartment).	Euromonitor^h
Wages	Average annual wages per capita in thousands of US dollars.	Market Line Advantage^d
Instrumental variables (external instruments)
GovExp	Total government expenditure as percentage of GDP.	Market Line Advantage^d
LftWng	Share of the population of country i in year t voting for (rather) left-wing parties.	Own research; Briglauer et al. (2022)
StateAid31 FiberCov31	Hausman-type geographic instruments constructed as described in Section 4.2.	Own research

^a All sources listed below last accessed February 5th 2023.

^b The FTTH Council Europe is a non-profit industry organization that provided data to FTTH Council Europe members in the past at: https://www.ftthcouncil.eu/

^c Data is publicly available at: https://data.oecd.org/broadband/fixed-broadband-subscriptions.htm#indicator-chart

^d Data is commercially available at: https://www.marketline.com/

^e Data is publicly available at: https://www.oecd-ilibrary.org/economics/gross-domestic-product-gdp/indicator/english_dc2f7aec-en

^f Data is publicly available at: https://data.worldbank.org/indicator/SP.POP.DPND

^g Data is publicly available at: https://www.itu.int/en/ITU-D/Statistics/Pages/stat/default.aspx

^h Data is commercially available at: https://www.euromonitor.com/our-expertise/passport

ⁱ Data is publicly available at: https://data.worldbank.org/indicator/SP.RUR.TOTL

^j Data is publicly available at: https://www.oecd-ilibrary.org/finance-and-investment/interest-rates/indicator-group/english_86b91cb3-en

Table A2. National NBN state aid programs in OECD member states imposed during 2002–2019.

Region/State^a	Source^b
European states
EU27	Data for all EU27 member states are available at the EC’s website (https://ec.europa.eu/competition/elojade/isef/index.cfm?clear=1&policy_area_id=3), the Official Journal of the European Union, Atene KOM (2014; 2021), WIK et al. (2020) and Feasey, Bourreau, and Nicolle (2018)
NOR	Government: https://www.regjeringen.no/en/dokumenter/meld.-st.−23-2012-2013/id718084/sec1; Store Norske Leksikon: https://snl.no/H%C3%98YKOM-programmet; Telecompaper.com: https://www.telecompaper.com/news/norwegian-govt-provides-nok-160-mln-in-broadband-subsidies–1005412; Berkman Center (2010)
GBR	WIK (2009); Berkman Center (2010); OECD (2018); European Commission: https://ec.europa.eu/competition/state_aid/cases/243212/243212_1387832_172c_1.pdf
Non-European states
AUS	WIK (2009); Analysis Mason (2010); Given (2010); Berkman Center (2010); Beltrán (2014); Howell and Sadowski (2018); Government: https://www.infrastructure.gov.au/media-technology-communications/internet/national-broadband-network/nbn-legislative-framework; https://www.aph.gov.au/About_Parliament/Parliamentary_Departments/Parliamentary_Library/pubs/rp/BudgetReview201314/NBN#_ftn10
CAN	Analysis Mason (2010); Berkman Center (2010); Financial Post: http://business.financialpost.com/fp-tech-desk/crtc-declares-high-speed-internet-a-basic-service-creates-750-million-fund; Government of Canada: https://ised-isde.canada.ca/site/connect-to-innovate/en/funding-bring-high-speed-internet-canadians-rural-and-remote-communities-overview; http://www.ic.gc.ca/eic/site/119.nsf/eng/home; https://fcm.ca/sites/default/files/documents/resources/report/broadband-access-rural-canada.pdf OECD (2018)
JPN	WIK (2009); Berkman Center (2010); OECD: http://www.oecd.org/internet/ieconomy/43404360.pdf; https://www.oecd.org/sti/ieconomy/1811750.pdf; Government: https://www.soumu.go.jp/menu_seisaku/ict/u-japan_en/new_pckg03_menu.html
KOR	WIK (2009); Analysis Mason (2010); Berkman Center (2010); Government: http://www.vus.sk/broadband/nbbs/kr_nbbs.pdf; ANACOM: https://www.anacom.pt/render.jsp?categoryId=340674
NZL	Government: https://www.beehive.govt.nz/release/ultra-fast-broadband-investment-proposal-finalised; https://ufb.org.nz/nz-govt-increases-internet-connectivity-investment-to-2b/; https://www.crowninfrastructure.govt.nz/ufb/what/; https://www.centreforpublicimpact.org/case-study/rural-broadband-initiative-in-new-zealand; Analysis Mason (2010); Berkman Center (2010); WIK (2009); Given (2010); Beltrán (2014); Howell and Sadowski (2018)
USA	WIK (2009); Analysis Mason (2010); Berkman Center (2010); NTIA: https://www2.ntia.doc.gov/; https://www2.ntia.doc.gov/SBDD; https://ntia.gov/category/grants; FCC: https://www.fcc.gov/general/national-broadband-plan; https://www.fcc.gov/general/connect-america-fund-caf; Whitacre and Gallardo (2020)

^a Three-digit-ISO 3166–1 country codes reported in column 1.

^b For Belgium (BEL), Chile (CHL), Iceland (ISL), Mexico (MEX), Switzerland (CHE) and Turkey (TUR) to the best of our knowledge no, or only minimal (i.e. under de minimis rules), public subsidies have been granted for NBN deployment. Israel, due to its relatively high density of even its peripheral areas, does not have a national broadband plan as such, but instead aims to reach policy objectives through universal broadband obligations. The latter do not provide for compensatory mechanism for the two designated operators (OECD 2018, 40). In Luxemburg, we consider NBN deployment by the 100% state owned incumbent POST Luxembourg as state aid. In addition, information was provided on personal request of the authors by public authorities and study authors. All links last accessed February 5th 2023.

Table A3. Further robustness checks.

Dep. Var.: FiberCov(t)	(1)	(2)	(3)	(4)	(5)	(6)	(7)
FiberCov(t−1)	0.692*** (0.078)	0.677*** (0.083)	0.648*** (0.091)	0.676*** (0.092)	0.673*** (0.100)	0.677*** (0.087)	0.678*** (0.082)
StateAid(t)	0.279** (0.112)	0.357*** (0.088)	0.293*** (0.111)	0.382*** (0.086)	0.358*** (0.087)	0.353*** (0.086)	0.361*** (0.087)
FixMobSub(t)	0.524*** (0.182)	0.462** (0.189)	0.557*** (0.203)	0.484** (0.199)	0.497** (0.198)	0.483*** (0.182)	0.491*** (0.184)
AgeDepRatio(t)	0.003 (0.007)	0.002 (0.007)	0.005 (0.007)	0.001 (0.009)	0.002 (0.007)	0.002 (0.007)	0.001 (0.006)
GDPpc(t)	0.002 (0.004)	0.003 (0.004)	0.003 (0.004)	0.003 (0.005)	0.004 (0.005)	0.004 (0.004)	0.004 (0.005)
RuralPop(t)	−2.422* (1.466)	−3.202** (1.279)	−2.158 (1.739)	−3.343** (1.305)	−3.129** (1.345)	−2.888** (1.314)	−3.163** (1.322)
IntRate(t)	−0.020** (0.008)	−0.020** (0.008)	−0.021** (0.008)	−0.018** (0.008)	−0.021** (0.009)	−0.021** (0.009)	−0.021** (0.009)
ShareDSL(t)	0.193 (0.146)
CommExp(t)		−0.030 (0.037)
IntUsers(t)			0.003 (0.003)
TeleServP(t)				−0.001 (0.001)
Netflix(t)					0.002 (0.043)
ShareApprt(t)						0.007 (0.015)
Wages(t)							−0.000 (0.003)
F-test (p-value)	0.000	0.000	0.000	0.000	0.000	0.000	0.000
AR(2) test (p-value)	0.759	0.624	0.705	0.718	0.620	0.628	0.618
Hansen test (p-value)	0.153	0.103	0.129	0.151	0.128	0.119	0.118
# of instruments	27	27	27	27	27	27	27
# of groups/clusters	32	32	32	30	32	32	32
# of observations	512	512	512	480	512	512	512

Notes: All regressions use a one-step GMM-diff. estimator (Arellano and Bond 1991) and extend the results from Table 1, column (5) by including additional explanatory variables, as listed in the table. Standard errors in parentheses are clustered at the group (i.e. country) level and robust to arbitrary forms of heteroscedasticity. Note that the inclusion of the variable TeleServP in column (4) reduces the number of countries and observations accordingly (Table 2).

*p < 0.10.

**p < 0.05.

***p < 0.01.

Table A4. First-stage results.

	(1)	(2)	(3)
Dep. Var.:	ΔFiberCov(t−1)	ΔStateAid(t)	ΔFixMobSub(t)
ΔAgeDepRatio(t)	0.008 (0.015)	0.003 (0.023)	0.010** (0.004)
ΔGDPpc(t)	0.011*** (0.003)	0.000 (0.007)	0.002 (0.002)
ΔRuralPop(t)	−1.941 (2.504)	0.843 (3.782)	−1.064 (0.763)
ΔIntRate(t)	0.004 (0.006)	0.008 (0.008)	−0.001 (0.001)
StateAid(t-2)	0.027 (0.024)	−0.097*** (0.015)	0.009** (0.004)
FixMobSub(t-2)	−0.037** (0.017)	0.093*** (0.034)	−0.093*** (0.011)
FiberCov(t-2)	−0.005 (0.012)	−0.020 (0.013)	0.015*** (0.004)
ΔStateAid31(t)	0.066 (0.123)	0.225 (0.235)	0.369*** (0.045)
ΔFiberCov31(t−1)	0.910*** (0.138)	0.303* (0.161)	0.313*** (0.038)
ΔLftWng(t)	0.001* (0.000)	−0.002 (0.002)	−0.000 (0.000)
ΔGovExp(t)	0.007** (0.003)	−0.000 (0.005)	0.004** (0.002)
F-test of excluded instruments (p-value)	0.000	0.000	0.000
Partial R^2 of excluded instruments	0.155	0.078	0.243
# of groups/clusters	32	32	32
# of observations	512	512	512

Notes: All first-stage regressions pertain to the model reported in Table 1, column (3). Standard errors in parentheses are clustered at the group (i.e. country) level and robust to arbitrary forms of heteroscedasticity.

*p < 0.10.

**p < 0.05.

***p < 0.01.

OECD. 2018. Bridging the Rural Digital Divide. OECD Digital Economy Papers No. 265, Paris.

 Google Scholar

OECD. 2011. National Broadband Plans. OECD Digital Economy Papers, No. 181, Paris.

 Google Scholar

OECD. 2021. Bridging Connectivity Divides. OECD Digital Economy Papers No. 315, Paris.

 Google Scholar

WIK, VVA, Squire Patton Boggs, and LE Europe (2020). The Role of State Aid for the Broadband Networks Rapid Deployment of in the EU. Report prepared for the European Commission, Directorate-General for Competition, Brussels.

 Google Scholar

Anderson, T. W., and C. Hsiao. 1981. “Estimation of Dynamic Models with Error Components.” Journal of American Statistical Association 76 (375): 598–606. https://doi.org/10.1080/01621459.1981.10477691.

 Web of Science ®Google Scholar

Arellano, M., and S. Bond. 1991. “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations.” Review of Economic Studies 58 (2): 277–297. https://doi.org/10.2307/2297968.

 Web of Science ®Google Scholar

Feasey, R., M. Bourreau, and A. Nicolle. 2018. State Aid for Broadband Infrastructure in Europe – Assessment and Policy Recommendations. CERRE Report. https://cerre.eu/publications/state-aid-broadband-infrastructure-assessment-policy-recommendations/.

 Google Scholar

Briglauer, W., C. Cambini, and M. Grajek. 2018. “Speeding Up the Internet: Regulation and Investment in the European Fiber Optic Infrastructure.” International Journal of Industrial Organization 61:613–652. https://doi.org/10.1016/j.ijindorg.2018.01.006.

 Web of Science ®Google Scholar

Briglauer, W., K. Gugler, C. Cambini, and V. Stocker. 2022. “Net Neutrality and High-Speed Broadband Networks: Evidence from OECD Countries.” European Journal of Law and Economics. https://doi.org/10.1007/s10657-022-09754-5.

 Web of Science ®Google Scholar

Atene KOM. 2014. Study on National Broadband Plans in the EU-28. Study Prepared for the European Commission (SMART Number 2014/077). Brussels: European Union.

 Google Scholar

Atene KOM. 2021. Study on National Broadband Plans in the EU-27. Study Prepared for the European Commission (SMART 2019/0007). Brussels: European Union.

 Google Scholar

Berkman Center. 2010. “Next Generation Connectivity: A Review of Broadband Internet Transitions and Policy from Around the World.” Final Report at Harvard University.

 Google Scholar

WIK. 2009. Breitband/Bandbreite für alle: Kosten und Finanzierung einer nationalen Infrastruktur. WIK-Diskussionsbeitrag Nr. 330, Bad Honnef.

 Google Scholar

Analysis Mason. 2010. Contemporary FTTH Stimulus Packages Study – v.3. Report for the FTTH Council Europe.

 Google Scholar

Given, J. 2010. “Take Your Partners: Public Private Interplay in Australian and New Zealand Plans for Next Generation Broadband.” Telecommunications Policy 34 (9): 540–549. https://doi.org/10.1016/j.telpol.2010.07.012.

 Web of Science ®Google Scholar

Beltrán, F. 2014. “Fibre-to-the-home, High-Speed and National Broadband Plans: Tales from Down Under.” Telecommunications Policy 38 (8–9): 715–729. https://doi.org/10.1016/j.telpol.2013.08.006.

 Web of Science ®Google Scholar

Howell, B., and B. Sadowski. 2018. “Anatomy of a Public-Private Partnership: Hold-up and Regulatory Commitment in Ultrafast Broadband.” Telecommunications Policy 42 (7): 552–565. https://doi.org/10.1016/j.telpol.2018.05.001.

 Web of Science ®Google Scholar

Whitacre, B., and R. Gallardo. 2020. “State Broadband Policy: Impacts on Availability.” Telecommunications Policy 44 (9): 102025. https://doi.org/10.1016/j.telpol.2020.102025.

 PubMed Web of Science ®Google Scholar